GENE DUPLICATION AND FUSION: 
STRATEGY FOR ACTIVE SITE CONTROL AND STARTING POINT 
FOR NEW CATALYSTS 
 
by 
 
Yu Wang 
 
 
 
 
A dissertation submitted to the Graduate Faculty of 
Auburn University 
in partial fulfillment of the 
requirements for the Degree of 
Doctor of Philosophy  
 
Auburn, Alabama 
Dec 8, 2012 
 
 
 
 
Keywords: catalase-peroxidase, KatG, folding platform, heme protein,  
C-terminal domain, gene duplication and fusion 
 
 
Copyright 2012 by Yu Wang 
 
 
Approved by 
 
Douglas Goodwin, Chair, Professor of Biochemistry 
Holly Ellis, Professor of Biochemistry 
Evert Duin, Professor of Biochemistry 
Christian Goldsmith, Professor of Chemistry 
 iv 
 
 
 
 
 
Abstract 
 
 Catalase?peroxidases (KatGs) have two peroxidase-like domains. The N-terminal domain 
contains the heme-dependent, bifunctional active site. Though the C-terminal domain lacks the 
ability to bind heme or directly catalyze any reaction, it has been proposed to serve as a platform 
to direct the folding of the N-terminal domain. Toward such a purpose, its I?-helix is highly 
conserved and appears at the interface between the two domains. Single and multiple substitution 
variants targeting highly conserved residues of the I?-helix were generated for intact KatG as well 
as the stand-alone C-terminal domain (KatG
C
). Single variants of intact KatG produced only 
subtle variations in spectroscopic and catalytic properties of the enzyme. However, the double 
and quadruple variants showed spectroscopic and catalytic properties similar to that observed for 
the N-terminal domain on its own (KatG
N
). The analogous variants of KatG
C
 showed a much 
more profound loss of function as evaluated by their ability to return KatG
N
 to its active 
conformation. These results suggest that the I?-helix is central to direct structural adjustments in 
the adjacent N-terminal domain. In particular, substitution of E695, a strictly conserved residue 
of the I?helix was especially destructive to KatG active site integrity. Available structures of 
KatG indicate that E695 is a central component of several hydrogen-bonded networks that also 
include strictly conserved R126 and W159 from the adjacent N-terminal domain. Both W159 (N-
terminus of the D-helix) and R126 (BC helical connecting loop) are part of core structures of the 
peroxidase-like N-terminal domain with connections to the active site. This points to a potential 
connection between the I?-helix and its influence on active site conformation and function. To 
 v 
evaluate this hypothesis, we replaced R126, W159 and E695 with alanine, singly and in 
combination, not only for intact KatG, but also for the stand-alone KatG
N 
or KatG
C
, as 
appropriate. Single variants of intact KatG showed a substantial loss of stability, particularly at 
the active site. The analogous variants of KatG
C 
and KatG
N
 showed a profound loss of function 
as evaluated by the return KatG
N
 (or its variants) to a functional conformation, suggesting C-
terminal domain through these interactions directs active site structural adjustments that are 
essential for catalytic function in the active site 25 ? away. The C-terminal domain not only 
directs conformational adjustments in its N-terminal domain partner, but due to its origin from 
gene duplication, also still retains the helical architecture of a typical peroxidase active site. This 
site lacks the residues necessary for catalytic activity or even heme binding. As such, the C-
terminal domain appears to provide an ideal ?blank slate? for engineering new heme-dependent 
catalysts. Spectroscopic measurements showed that a M616G/R617H variant of KatG
C
 was 
sufficient to restore heme binding, producing a hexacoordinate low-spin ferric state.  Additional 
modifications produced spectra very similar to those observed for KatG
N
.  
 
 
 
 
 
 
 
 
 vi 
 
 
 
 
 
Acknowledgments 
 
 The research conducted in this dissertation would not have been possible without the help 
and support of many people whom I wish to thank. Listed here are only a few of those. First, I 
would like to thank my advisor Dr. Douglas Goodwin. He guided me through my whole 
research, trained me to be an independent researcher. He had a significant influence on my future 
career and personal life. To me he is more than an advisor; without his encouragement, I could 
not have reached today?s achievement. I wish I will continue to make him proud.  Second, I 
would like to thank Dr. Holly Ellis for taking the time to review every manuscript I have written. 
Dr. Evert Duin has been of great assistance with EPR. Third, I would like thank all my co-
workers at Auburn University, especially Robert Moore and Carma Cook. They provided great 
help and assistance when I first started my research. I also feel lucky to have worked with 
Shalley Kudalkar, Ellizebeth Ngwane and Haijun Duan. Especially I thank my family for their 
continuous understanding, help, support and encouragement for these five years. Finally, I would 
like to thank NSF, the Department of Chemistry and Biochemistry, Auburn University?s 
graduate school, and COSAM for their funding support.  
 
 
 
 
 
 vii 
 
 
 
 
 
Table of Contents 
 
Abstract ......................................................................................................................................... iv 
Acknowledgments ......................................................................................................................... vi 
CHAPTER ONE ............................................................................................................................ 1 
1.1 Hydrogen peroxide and its scavengers in cels ...................................................................................... 2 
1.2 Typical (i.e., monofunctional) catalases ................................................................................................ 3 
1.3 Peroxidases ............................................................................................................................................. 6 
1.4 Catalase-peroxidases (KatGs) .............................................................................................................. 40 
CHAPTER TWO ......................................................................................................................... 54 
2.1 Introduction .......................................................................................................................................... 54 
2.2 Materials and methods ......................................................................................................................... 57 
2.3 Results .................................................................................................................................................. 66 
2.4 Discusions ........................................................................................................................................... 85 
CHAPTER THREE ..................................................................................................................... 94 
3.1 Introduction .......................................................................................................................................... 94 
3.2 Materials and methods ......................................................................................................................... 96 
3.3 Results ................................................................................................................................................ 102 
3.5 Discusion .......................................................................................................................................... 121 
CHAPTER FOUR ...................................................................................................................... 134 
4.1 Introduction ........................................................................................................................................ 134 
 viii 
4.2 Materials and methods ....................................................................................................................... 138 
4.3 Results ................................................................................................................................................ 141 
4.4 Discusion and conclusion ................................................................................................................. 152 
CHAPTER FIVE ....................................................................................................................... 157 
5.1 Introduction ........................................................................................................................................ 157 
5.2 Methods .............................................................................................................................................. 159 
5.3 Results ................................................................................................................................................ 160 
5.4 Conclusion ......................................................................................................................................... 165 
CHAPTER SIX .......................................................................................................................... 167 
REFERENCES .......................................................................................................................... 174 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ix 
 
 
 
 
 
List of Figures 
 
Figure 1.1: Overall structure (A), one subunit (B) and the active site (C) of the E.coli catalase 
(HP II) ............................................................................................................................5?
Figure 1.2: Catalytic mechanism of typical catalases ......................................................................7?
Figure 1.3 Overview of the classification of heme peroxidases (super) families ..........................10?
Figure 1.4: Catalytic cycle of bacterial diheme cytochrome c peroxidase ....................................12?
Figure 1.5:  Structure of di-heme-cytochrome c peroxidase and conformational change during 
catalysis ........................................................................................................................14?
Figure 1.6 Structures of DyP and DyP-type peroxidase ................................................................16?
Figure 1.7: Schematic diagram of the proposed formation of compound I by DyP ......................18?
Figure 1.8: A modified mechanism of compound I formation by DyP .........................................18?
Figure 1.9: Structure of CPO .........................................................................................................20?
Figure 1.10: Catalytic cycle of haloperoxidases ............................................................................22?
Figure 1.11: Proposed mechanism of compound I formation by chloroperoxidase ......................23?
Figure 1.12: Structure of human MPO ..........................................................................................26?
Figure 1.13: Catalytic cycles of MPO ............................................................................................27?
Figure 1.14: Structure of HRP .......................................................................................................31?
Figure 1.15: Peroxidase cycle of HRP ...........................................................................................32?
Figure 1.16: Structure of Class II peroxidases. ..............................................................................36?
Figure 1.17: Catalytic cycle of manganese peroxidase (MnP) ......................................................37?
 x 
Figure 1.18: Structure of Class I peroxidase. .................................................................................41?
Figure 1.19: Comparisons of the overall structures and the active sites of monofunctional 
peroxidases, KatG and monofunctional catalases. .......................................................44?
Figure 1.20: Two novel disulfide bridges between the N-terminal domains of two subunits in 
KatG .............................................................................................................................46?
Figure 1.21: Structure of catalase-peroxidase. ...............................................................................48?
Figure 1.22: The covalent adduct in KatG. ....................................................................................49?
Figure 1.23: Classical model of catalytic cycle of KatG (A), modified model of catalase 
mechanism of KatG (B). ..............................................................................................50?
Figure 2.1: (A) Far-UV circular dichroism of wtKatG, and I?-helix variants. (B) Far-UV circular 
dichroism of C-terminal domain I?helix variants and KatG
C
 ......................................67?
Figure 2.2: Absorption spectra of ferric wtKatG, its I'-helix substitution variants, and KatG
N
 ....68?
Figure 2.3: MCD spectra of ferrous wtKatG, KatG
N 
and I?-helix substitution variants. ...............70?
Figure 2.4: EPR spectra recorded for wtKatG, its I'-helix substitution variants, and KatG
N
. .......71?
Figure 2.5: GuHCl-mediated unfolding of wtKatG, its I'-helix variants, and KatG
N
 as monitored 
by far-UV CD spectroscopy .........................................................................................78?
Figure 2.6: Thermal stability of heme cavity of wtKatG and its I'-helix variants .........................79 
Figure 2.7: UV?visible absorption spectra of KatG
N
 following incubation with KatG
C
 or its I'-
helix substitution variants ............................................................................................81?
Figure 2.8: EPR spectra of KatG
N
 following incubation with KatG
C
 or its I'-helix substitution 
variants .........................................................................................................................83?
Figure 2.9: Time course for recovery of KatG
N
 catalase activity upon addition of equimolar 
KatG
C
 or one of its I'-helix substitution variants .........................................................86?
 xi 
Figure 2.10: Interfaces between the KatG N-terminal domain and I'-helix from the C-terminal 
domain ..........................................................................................................................92?
Figure 3.1: Far-UV circular dichroism of wtKatG, and its E695, R126 and W159 substitution 
variants .......................................................................................................................103?
Figure 3.2: UV-visible absorption spectra of ferric wtKatG and its E695, R126 and W159 
substitution variants ...................................................................................................104?
Figure 3.3: MCD spectra of ferrous wtKatG and its E695A, R126A and W159A .....................106 
Figure 3.4: EPR spectra of wtKatG
 
and its E695, R126 and W159 substitution variants ...........107?
Figure 3.5: GuHCl-mediated unfolding of wtKatG and E695, R126 and W159 substitution 
variants as monitored by far-UV circular dichroism (CD) spectroscopy. .................113?
Figure 3.6: Thermal stability of the heme cavity of wtKatG, E695, R126 and W159 substitution 
variants as monitored by UV-visible absorption due to the heme retention ..............115?
Figure 3.7: UV-visible absorption spectra of KatG
N 
(or its R126A, W159A substitution variants) 
following incubation with KatG
C
 (or its E695A, E695Q substitution variants). .......118?
Figure 3.8: EPR spectra of KatG
N 
(or its R126A, W159A substitution variants) following 
incubation with KatG
C
 (or its E695A, E695Q substitution variants) ........................119?
Figure 3.9: Time course for recovery of KatG
N
 (or its R126A, W159A)
 
catalase activity upon 
addition of equimolar KatG
C
 (or its E695A, E695Q) ................................................124 
Figure 3.10: Interactions between the N-terminal domain BC loop and the B?C? loop from the 
intrasubunit C-terminal domain .................................................................................127?
Figure 3.11: Interactions between R126, W159 and E695 and water-mediated hydrogen-bonded 
networks along BC loop; BC loop in manganese peroxidase and Ca
2+ 
 ....................130 
Figure 3.12: Overlaid structures of KatGs and the water molecules ...........................................131?
 xii 
Figure 4.1: Overlaid structures of KatG
C
 (green) and CCP (cyan) with a bound heme (red). ....136?
Figure 4.2: Comparison of the ?active site? environments of APX and KatG
C
 ..........................137?
Figure 4.3: Far-UV circular dichroism of KatG
C
 and KatG
C
 variants. ........................................144?
Figure 4.4: UV-vis spectra of ferric KatG
C
, KatG
N 
and KatG
C
 variants (HB 1) .........................146?
Figure 4.5: UV-vis spectra of ferric KatG
C
, KatG
N 
and KatG
C
 variants .....................................147?
Figure 4.6: Magnetic circular dichroism (MCD) spectra for KatG
C
 and KatG
N
 variants ...........150?
Figure 4.7: EPR spectra for ferric state of KatG
C
 and KatG
N
 variants. .......................................151?
Figure 5.1: Distal active site M-Y-W covalent adduct in KatG from M. tuberculosis ................158?
Figure 5.2: Evaluation of M-Y-W covalent adduct formation by SDS-PAGE ...........................161?
Figure 5.3: The influence of peripheral structures on the formation of M-Y-W adduct. ............163?
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 xiii 
 
 
 
 
 
List of Tables 
 
Table 2.1: Primer sets for site-directed mutagenesis of I?-helix. ...................................................59?
Table 2.2: Ratios of EPR signals observed for wtKatG, its I'-helix variants, and KatG
N
. ............73?
Table 2.3: Apparent catalase kinetic parameters for wtKatG and its I'-helix variants  .................74?
Table 2.4: Apparent peroxidase kinetic parameters for wtKatG and its I?-helix variants ..............75?
Table 2.5: Parameters GuHCl-mediated unfolding of wtKatG and I?-helix variants ....................77?
Table 2.6: Ratios of EPR signals observed following KatG
N
 incubation with KatG
C
 or its I'-helix 
substitution variants .......................................................................................................84?
Table 2.7: KatG
N
 reactivation k
obs
 and apparent catalase parameters for reactivated enzyme. .....87?
Table 2.8: Apparent Peroxidase parameters for reactivated KatG
N
 by KatG
C
 or its I?-helix 
variants ..........................................................................................................................88?
Table 3.1: Primer sets for site-directed mutagenesis .....................................................................98?
Table 3.2: Ratios of EPR signals observed in wtKatG and its E695, R126 and W159 substitution 
variants. .......................................................................................................................109?
Table 3.3: Apparent catalase kinetic parameters for wtKatG and its E695, R126 and W159 
substitution variants
 
 ....................................................................................................110?
Table 3.4: Apparent peroxidase kinetic parameters for wtKatG and its E695, R126 and W159 
substitution variants
 
 ....................................................................................................111?
Table 3.5: Parameters of GuHCl-mediated unfolding wtKatG and E695, R126 and W159 
substitution variants. ....................................................................................................114?
 xiv 
Table 3.6: Ratios of EPR signals observed following KatG
N
 (or its R126A, W159A) following 
incubation with KatG
C
 (or its E695A, E695Q) ...........................................................120?
Table 3.7: Apparent catalase kinetic parameters for KatG
N
 (or its R126A, W159A) following 
incubation with KatG
C
 (or its E695A, E695Q)
 
 ..........................................................122?
Table 3.8: Apparent peroxidase kinetic parameters for KatG
N
 (or its R126A, W159A) following 
incubation with KatG
C
 (or its E695A, E695Q substitution variants)
 
 .........................123?
Table 4.1: Primer sets for site-directed mutagenesis of KatG
C
 ...................................................140?
Table 4.2: Variants name corresponding to the detailed mutagenesis strategy ...........................143?
Table 4.3: Absorption characteristics of KatG
N
 and KatG
C
 variants ...........................................148?
 1 
 
 
 
 
 
CHAPTER ONE 
 
 
LITERATURE REVIEW 
 
 
The central proposes of the research described in this dissertation are to evaluate the role of 
the C-terminal domain in catalase-peroxidase (KatG) enzymes, and to borrow this structure as a 
template for engineering new heme-dependent catalysts. The two-domain structure of KatG is a 
novel feature in comparison to other members of its superfamily as is its bifunctional catalytic 
capability. The connection between KatG?s structure and its function remains poorly defined, but 
its contrast against other structurally similar but functionally distinct relatives, provides an 
excellent opportunity to evaluate the mechanism by which structures peripheral to and distant 
from an active site influence structure and function.  More than this, it affords an ideal system to 
apply the information and strategy toward engineering new heme enzymes. This chapter will first 
establish the importance of catalases and peroxidases in nature, and then discuss the catalytic 
mechanisms employed by these H
2
O
2
-degrading enzymes by comparing their structures and 
functions in order to understand the importance of the two-domain structure, bifunctionality and 
molecular phylogeny of the catalase-peroxidases. Furthermore, discussion focused on the 
structure and function relationships of various heme-dependent peroxidases will shed light on the 
structural basis and strategy for engineering the KatG C-terminal domain to a new heme-
dependent catalyst, building on its unique origin and features. 
 
 2 
1.1 Hydrogen peroxide and its scavengers in cells: 
Hydrogen peroxide is one of the most common reactive oxygen species that aerobic 
organisms encounter. For example, in well-fed aerobic E.coli cells, H
2
O
2
 is formed at a constant 
rate of 10-15 ?M/s [1]. This endogenous H
2
O
2
 is primarily formed from reduction of O
2
 by 
flavoproteins [2]. Exogenous H
2
O
2
 derived from environmental sources is also an issue because 
H
2
O
2
 can penetrate cell membranes. Exogenous sources of H
2
O
2
 include the NADPH oxidase-
based response of plants and animials to infection, H
2
O
2
-excreting microbes, extracellular 
oxidation at anoxic/oxic interfaces, and photochemically-driven oxidations [2]. Due to its ability 
to initiate damage to biomolecules, excessive H
2
O
2
 is toxic for living cells. In particular, it can 
destroy iron-sulfur proteins involved in energy production and biosynthesis [3], and deactivate 
other enzymes that rely on ferrous iron as a cofactor [4]. The reduction of H
2
O
2
 by ferrous iron 
produces the highly reactive hydroxyl radical (?OH) by the well-known Fenton reaction. The 
unregulated generation of such a reactive oxidant is known to result in damage to all classes of 
biological molecules. Therefore, rapid and efficient removal of excess H
2
O
2 
is essential for all 
organisms that are to survive in an aerobic enviornment. In most organisms, H
2
O
2 
is scavenged 
by peroxidases and/or catalases. The latter show a striking ability to disproportionate H
2
O
2
 to 
water concomitant with the generation of dioxygen, according to reaction (1). Enzymes with 
catalase activity are predominantly found in three gene families: typical (i.e., monofunctional) 
catalases, catalase-peroxidases (KatGs), and manganese (non-heme) catalases [5]. Among these, 
the large majority are heme-containing enzymes, and these are widely distributed among 
prokaryotes and eukaryotes. Manganese catalase is a minor gene group only present in bacteria; 
they are less proficient catalysts than typical catalases and catalase-peroxidases [5]. Though 
important, the manganese catalases are not particularly relevant to this discussion and will not be 
 3 
covered further. 
 
 
1.2 Typical (i.e., monofunctional) catalases: 
Phylogenetic analyses indicate that there are three main evolutionary clades, which were 
segregated in an early stage of evolution by way of several gene-duplication events [5, 6]. Clade 
1 catalases are small-subunit proteins with heme b as the prosthetic group. They are distributed 
among bacteria, algae and plants. Clade 2 contains the large-subunit catalases, and these 
primarily use heme d as prosthetic group and contain an additional ?flavodoxin-like? domain. 
These are widely distributed among bacteria and fungi. Clade 3 is the most widely distributed; 
being abundant in archaebacteria, fungi, protists, plants and animals. These are typically small-
subunit catalases that contain heme b and NADPH as cofactors. 
 
1.2.1 Structure and active site of monofunctional catalases: 
The monofunctional catalase from E. coli, hydroperoxidase II (HPII), is a tetrameric large-
subunit catalase (Fig. 1.1A) found in clade 2 of the typical catalase family. Available crystal 
structures indicate that HP II is composed of five different tertiary structural regions or domains, 
including 1) the amino terminal arm, 2) an antiparallel eight-stranded ?-barrel, 3) an extended 
wrapping loop, 4) a helical domain, and 5) a C-terminal, flavodoxin-like domain [7]. The core of 
the protein is similar in structure and sequence to those of small subunit catalases; however the 
75-80 residues of the amino-terminal arm and the flavodoxin-like domain are unique to HPII and 
other clade 2 enzymes (Fig. 1.1B). It has been proposed that both structural elements are 
dedicated to facilitating the folding and oligomerization process [7, 8].  
 4 
Although significant differences exist in the overall folding between small- and large-
subunit catalases, the active site residues are strictly conserved among all three clades (Fig. 1.1C) 
[5]. These include Asn 201 and His 128 (HPII numbering) in the distal site, and Tyr 415, which 
ligates the iron on the heme?s proximal side. Asn 201 is proposed to stabilize the peroxide [9]. 
The imidazole ring of His 128 is situated 3.5 ? above and parallel to the plane of the heme, 
allowing for an H-bond with a conserved Ser 167 [7, 10]. The coplanar position of the imidazole 
ring contrasts with the perpendicular orientation of this group in peroxidases. This is proposed to 
allow the second oxygen of the peroxide to form an H-bond with the imidazole during catalysis, 
facilitating O
2
 production [5]. In all monofunctional catalases, a tyrosine has been identified as 
the proximal heme iron ligand in contrast to a histidine in plant peroxidases. The strong anionic 
character of the ligand is expected to favor the oxidized heme species [11]. Interestingly, a 
unique feature for E. coli HPII is a covalent bond between the proximal Tyr ligand and an 
adjacent imidazole ring from His 392 [12]. This has been proposed to enhance resistance to 
peroxide-dependent deactivation by preventing formation of compound II, an intermediate that is 
not part of the enzyme?s catalytic cycle [5]. 
 
Another feature typical of monofunctional catalases is a long, narrow substrate access 
channel leading to the deeply buried active site heme group [7]. In contrast, typical peroxidases 
usually provide more open access to the heme cofactor. In E.coli HPII, two major channels can 
be defined: one has small diameter and is full of hydrophobic residues, leading H
2
O
2
 to the heme 
distal pocket; the other is a large funnel-shaped opening positioned near residues 590 and 595 
and leads to the vicinity of Asn 201 [7]. The shape and size of these channels are essential for 
selectively directing substrates, particularly H
2
O
2
, to the active site [7, 13].  
 5 
 
 
 
Figure 1.1: Overall structure (A), one subunit (B) and the active site (C) of the E.coli 
catalase (HP II) (PDB: 1GGE). 
 
 
 
 
 
 
 
 6 
 
1.2.2 The catalytic mechanism of typical catalases: 
The catalase cycle starts with the oxidation of the heme iron along with the heterolytic 
cleavage of the O-O bond of hydrogen peroxide.  This process produces a ferryl-oxo porphyrin 
?-cation radical ([Por 
. 
+] Fe
IV
=O) intermediate known as compound I with the concomitant 
production of a water molecule (Fig. 1.2) [9]. Kinetics studies [14] indicate that once compound 
I forms, it is rapidly and efficiently reduced back to the resting (i.e., ferric) enzyme by reacting 
with a second equivalent of H
2
O
2
, generating H
2
O and O
2
 [15]. Two mechanisms have been 
proposed for O
2
 production and the return of the ferric state: 1) a His-mediated mechanism, or 2) 
a direct mechanism involving radical formation (Fig. 1.2) [16].  The radical mechanism posits a 
radical intermediate generated upon hydrogen atom transfer from H
2
O
2
 to the ferryl heme 
intermediate. This bears some resemblance to the mechanism proposed for alkane hydroxylation 
by cytochrome P450 [17]. In contrast, the His-mediated mechanism shows the participation of 
ionic species resulting from an initial proton abstraction carried out with the assistance of the 
distal His as the general base. A recent study of the catalase mechanism by QM/MM Car-
Parrinello Metadynamics simulations highlight the key role of distal residues acting as an 
acid/base catalyst [15], lending credence to the His-mediated mechanism. 
 
1.3 Peroxidases: 
The first part of this section will be dedicated to the structure and function relationships of 
heme dependent peroxidases from all five different superfamilies/families where the activity is 
 7 
 
 
Figure 1.2: Catalytic mechanism of typical catalases. 
 
 
 
 
 
 
 8 
found. Understanding these will provide a structural backdrop for understanding both the 
diversity and the commonalities of mechanisms and strategies by which heme-dependent 
peroxidase activity is accomplished. Because catalase-peroxidases (KatGs) fall into the 
peroxidase-catalase superfamily, the second part will progressively introduce members in this 
superfamily by discussing their phylogenic relationships and comparing their structures and 
functions with KatGs in order to understand the importance of the unique two-domain structure 
observed in KatGs.  Together, these sections will lay the groundwork for evaluating the 
mechanisms by which the C-terminal domain modulates the conformation and performance of an 
active site that is sequestered in the N-terminal domain ? 30 ? away.  They will also highlight 
strategies for using the stand-alone C-terminal domain as a scaffold for engineering new heme-
dependent catalysts. 
 
1.3.1 Systematic Classification of Peroxidases: 
Peroxidases are capable of reducing H
2
O
2 
to water with the concomitant one- or two-
electron oxidation of various organic and inorganic substrates according to reactions (2) and (3) 
[18]. Currently, 15 different Enzyme Commission (EC) numbers from EC 1.11.1.1 to EC 
1.11.1.16 have been assigned to peroxidases [19]. Although there are heme and non-heme 
examples of peroxidases, the most abundant are those containing heme as the prosthetic group. 
As of January 2012 PeroxiBase (a database dedicated to cataloging genomic and other data on 
peroxidases), nearly 75% of all genes encoding enzymes with peroxidase activity are the heme-
containing variety [20]. Traditionally, heme peroxidases have been divided into ?plant? and 
?animal? groups, but the influx of new of genomic data has rendered such schemes obsolete. 
According to the current scheme, available heme peroxidase gene sequences are grouped 
 9 
phylogenetically into three families (the di-heme peroxidases, the DyP-type peroxidases and the 
heme-haloperoxidases) and two superfamilies (the peroxidase-cyclooxygenase superfamily and 
the peroxidase-catalase superfamily) (Fig. 1.3) [21]. The majority of sequences fall into one of 
the two superfamilies, and the remaining sequences are distributed among the three families. 
 
 
1.3.2 Di-heme peroxidases family: 
1.3.2.1 Structure and function of di-heme peroxidases: 
Representatives of this family are found predominantly in bacteria with some members 
found in achaea. A detailed phylogeny has been proposed based on sequence data available with 
most of the sequences derived from the di-heme cytochrome c peroxidases (DiHCcP) [21]. As 
the name suggests these enzymes contain two hemes. They are typically periplasmic where they 
catalyze reduction of hydrogen peroxide to water using single electron donors such as 
cytochrome c or cupredoxin [22]. Though DiHCcPs bear no sequential or topological 
resemblance to KatGs, all of them have been shown to be dimeric where each subunit (35-40 
kDa) is composed of two domains [23], similar to the two-domain structure of KatGs. In contrast 
to KatGs where only one domain (the N-terminal domain) is catalytic active, both domains of 
DiHCcPs participate in catalytic events. There is an electron-transferring heme (E heme) domain 
and a peroxidatic heme (P heme) domain [23]. Both hemes are of the c-type in that they are 
covalently linked to the polypeptide chain by way of cysteine side chains [24]. In contrast, the  
 
 10 
 
 
 
Figure 1.3: Overview of the classification of heme peroxidases (super) families. 
 
 
 
 
 
 
 
 
 11 
canonical eukaryotic cytochrome c peroxidase (from the peroxidase-catalase superfamily) 
contains a single b-type heme in its active site. 
 
The P heme is a low-potential center where H
2
O
2
 is reduced. The E heme is a high-
potential center that serves to transfer electrons from soluble electron-shuttle proteins such as 
cytochrome c or azurin to the peroxidatic site [25-27]. The P heme is coordinated by two axial 
histidine ligands, whereas the E heme is ligated by one His and one Met residue [23, 24, 27-31]. 
Crystal structures of DiHCcP from Pseudomonas aeruginosa reveal that a single Ca
2+
 ion is 
tightly bound at the domain interface of both fully oxidized and mixed-valence forms of the 
enzyme. This Ca
2+
 is absolutely required for catalytic activity [23]. The role of the Ca
2+
 ion is 
unresolved, but it is proposed to maintain the structural integrity of the enzyme and/or modulate 
electron transfer between the two domains [23, 24, 27-31].  
 
1.3.2.2 Catalytic cycle and heme function: 
In most known bacterial DiHCcP enzymes, the inactive resting state can be converted to 
the active mixed-valence state by reduction of the E heme [32]. A mechanism proposed for the 
enzyme is shown in Fig. 1.4 [25]. The enzyme is isolated in an inactive state (state I) where both 
hemes are in a hexacoordinate ferric form. An electron (from ascorbate or the enzyme?s 
physiological substrates) is transferred to the E heme (step A). E heme reduction triggers a Ca
2+
-
dependent conformational change (step B) that results in the removal of the distal histidine from 
the iron coordination sphere where it is replaced by a water molecule [26, 32, 33]. This produces 
the active form of the enzyme (state III). State III is then oxidized by H
2
O
2
 at the P heme 
producing a ferryl intermediate and a ferric
 
E heme. This would be roughly analogous to the  
 12 
 
Figure 1.4: Catalytic cycle of bacterial diheme cytochrome c peroxidase. 
 
 
 
 
 
 
 
 
 
 
 
 
 13 
compound I state of the monoheme peroxidases of eukaryotes (step C). The catalytic cycle is 
completed by the delivery of two electrons in single-electron steps (steps D and E) from 
cytochrome c or other substrates. Stage V is electronically equivalent to the resting enzyme, but 
its P heme is accessible to ligands. Relaxation back to the resting state is likely to be a slow 
process. It has been proposed that this step is only observed when insufficient reductant is 
present [32]. 
 
1.3.2.3 Enzyme structure and conformational changes during catalytic cycle: 
As described earlier, the bacterial DiHCcP is composed of two heme-containing domains 
[23, 24, 27-31]. Both oxidized and mixed-valence forms of the enzyme share similar topological 
structures (Fig. 1.5). In the E heme-containing domain, the coordination of the heme remains 
unchanged by oxidation state. Conversely, the changes of the orientation of heme D-ring 
propionate group leads to a loss of an H-bond with the Leu 216, and in the mixed-valence state, 
Leu 216 moves away from the proximal heme site. The A-ring propionate group linked to Ca
2+
 
ion does not change under activation. In the P heme-containing domain, more dramatic changes 
occur during activation. The loop carrying His 71 moves toward the dimer interface concomitant 
with relocation of the aromatic side chain of Trp 73. The movement of this loop allows for the 
occupation of peroxidatic site by a water molecule, and the introduction of two other conserved 
residues Gln 104 and Glu 114, which have been suggested to play critical roles in the formation 
and stabilization of the oxy-ferryl intermediate [23, 27]. 
 14 
 
E heme domain P heme domain 
His 201 
Met 275 
D 
A 
Leu 216 
Leu 216 
Ca
2+ 
Ca
2+ 
Ca
2+ 
Ca
2+ 
His 71 
His 55 
His 71 
His 55 
Arg 97 
Arg 97 
Glu 114 
Glu 114 
Trp 73 
Trp 73 
Gln 104 
Gln 104 
A B 
C 
D 
 
 
Figure 1.5:  Structure of di-heme-cytochrome c peroxidase and conformational changes 
during catalysis. 
The overall structure of di-heme cytochrome c peroxidase from Pseudomonas aeruginosa (A); 
the conformational change in the E heme domain (B); the oxidized state of P heme domain (C); 
the mixed-valence state of P heme domain (D). The spheres represent Ca
2+
. The protein structure 
colored by green and heme highlighted by red represents the oxidized state (PDB: 1eb7) [27]; the 
protein structure colored by canyon and heme highlighted by hot pink represents the mixed-
valance state (PDB: 2vhd) [27].   
 
 15 
1.3.3 DyP-type peroxidase family: 
Dye-decolorizing peroxidase (DyP) was discovered in the basidiomycete Thanatephorus 
cucumeris, and it has been shown that this enzyme has the ability to decolorize 18 types of 
reactive, acidic and dispersive dyes, most of which are xenobiotics [34]. It is a member of a 
newly discovered and intriguing family of heme peroxidases, the DyP-type peroxidase family. 
So far, these enzymes have been found only in bacteria and fungi [35]. In this family, four 
distinct subfamilies (from A to D) can be identified, and it has been surmised that through 
evolution, subfamilies A and B first diverged from C and D; subsequently, produced the four 
noted subfamilies [21]. DyP is a member of subfamily D, whereas TyrA, which we will 
introduce later, belongs to subfamily B. 
 
1.3.3.1 Structures of DyP-type peroxidases: 
Consistent with its identity as a peroxidase, DyP contains a heme cofactor and requires 
H
2
O
2
 for all reactions. However, DyP has several characteristics that distinguish it from other 
typical peroxidases. The level of sequence homology with other peroxidases is very low. It is 
capable of using a wide range of substrates and is tuned to operate at a surprisingly low pH 
(optimum 3 - 3.2) [36, 37]. Initially, DyP was considered to be a class II member of the 
peroxidase-catalase superfamily, however, it became increasingly apparent that the primary 
structure was distinct. Indeed, DyP even lackes the active site signature motif R-X-X-F/W-H that 
is a hallmark of the peroxidase-catalase superfamily [38]. However, discussion of their structure 
and function points toward a strategy that might be employed for engineering the C-terminal 
domain of KatG into a new low pH peroxidase. 
 
 16 
 
 
Figure 1.6: Structures of DyP and DyP-type Peroxidase. 
A: Overall structure of DyP from Thanatephorus cucumeris Dec 1 (PDB: 2D3Q) [39]; B: 
Overall structure of TyrA from Shewanella oneidensis (PDB: 2IIZ) [40]; C: The heme and its 
surrounding residues important for the peroxidase activity of DyP; D: The active site 
surrounding of TyrA. 
 17 
A striking distinction in active site structure is the absence of the typical distal His, instead, 
its replacement by an aspartate (Asp 171) [35]. Additionally, DyP contains a unique pair of anti-
parallel ?-sheets located above the distal side of the heme (Fig. 1.6 A) [35]. Although sequence 
homology between DyP and other DyP-type peroxidase is very low, their overall tertiary 
structures are quite similar. TyrA, a member of type B DyP-type peroxidases, is a two-domain 
? + ? protein. Each domain contains a four-stranded, antiparallel ?-sheet sandwiched by ?-
helices in a ferredoxin-like fold (Fig. 1.6 B) [40]. At the active site level, three residues: an Asp, 
an Arg and the proximal His ligand are strictly conserved in this family (Fig 1.6 C and D). The 
Arg is proposed to assist with the formation of compound I by stabilizing higher oxidation states 
of heme?s iron center [35]. Asp instead of His as a general acid/base catalyst provides an 
explanation for the extraordinarily low pH optimum of DyP and DyP-type peroxidases [36, 37]. 
 
1.3.3.2 Catalytic cycle of DyP and Dyp-type peroxidases: 
A mechanism for the formation of compound I based on the participation of Asp and Arg 
in the active site has been proposed (Fig. 1.7) [35]. According to this model, a conserved Glu 
391 forms an H- bond with proximal ligand His 308, increasing the anionic character of the 
ligand, serving to stabilize higher heme iron oxidation states. In the resting state of DyP, a sixth 
ligand to the heme iron is absent, allowing for direct formation of a complex with H
2
O
2
 upon its 
entrance to the active site. Asp 171 acts as a general base to abstract a proton from the peroxide 
and deliver it to the distal peroxide oxygen atom concomitant with heterolytic cleavage of the O?
O bond. Subsequently, one electron is removed from iron, and another electron is transferred 
from porphyrin to the remaining oxygen, generating an intermediate analogous to compound I 
observed in other peroxidases. Recently, a modified mechanism for formation of compound I has  
 18 
 
Figure 1.7: Schematic diagram of the proposed formation of compound I by DyP. 
 
Figure 1.8: A modified mechanism of compound I formation by DyP. 
 19 
been proposed [39] (Fig. 1.8). In this model, the carboxylate group of Asp 171 associates with 
the ligand for heme iron by swinging into the appropriate position [39]. Once the compound I is 
formed, the carboxylate group can revert to its original position in order to stabilize compound I 
through an H-bonded network involving an active site water molecule. Interestingly, a compound 
II (ferryl) has never been detected in these enzymes [35]. To date, the picture of the catalytic 
mechanism of Dyp (or DyP-type) peroxidases remains incomplete.   
 
1.3.4 Haloperoxidases Family: 
Haloperoxidase activity has been observed with a variety of heme- or vanadium-dependent 
enzymes. However, the non-heme (vanadium) haloperoxidases are phylogenetically unrelated to 
this family. The heme-dependent haloperoxidases are abundant among fungi. They contain 
protoporphyrin IX as a prosthetic group and catalyze the oxidative transformation of halides and 
halophenols [41]. A recently constructed phylogenetic tree for this family indicates that heme 
haloperoxidases form a monophyletic group with frequent gene duplication events [21]. 
 
1.3.4.1 Structure and catalytic mechanisms of haloperoxidases family: 
In this family, the most intensively investigated member is the chloroperoxidases (CPO) 
from the ascomycete Caldariomyces fumago. This CPO is a 42 kDa glycoprotein. It catalyzes the 
H
2
O
2
-dependent chlorination of cyclopentanedione as well as the iodination and bromination of a 
wide range of substrates [42-46].  
 
 
 20 
 
Figure 1.9: Structure of CPO. 
Overall structure of CPO with binding of bromide (A) (PDB: 2CIV) and iodide (B) (PDB: 
2CIW) [47], and the active site of CPO (C). Dots represent the halides. 
 
 
 
 
 
 21 
Available structures from the protein data bank indicate that CPO has a novel tertiary 
structure dominated by eight helical segments [48]. In the active site, a cysteine serves as the 
proximal ligand to the heme iron. Unlike other peroxidases, there is no Arg in the distal cavity, 
and the catalytic acid/base required to cleave the peroxide O-O bond is Glu rather than His 
(Fig.1.9).  Consequently, CPO is active at low pH with an optimum at 2.8 [42, 47]. This unique 
active site configuration points to a future strategy by which our stand-alone KatG C-terminal 
domain (KatG
C
) may be engineered. Crystal structures solved in the presence of bromide and 
iodide have been reported. The bromide/iodide binding sites have been identified. The location 
of these ions points toward a route by which halides may access the heme center [47]. 
 
1.3.4.2 Catalytic cycle of CPO: 
The fundamental mechanism of CPO-catalyzed halogenations is presented in Fig 1.10 [41]. 
It begins with the heterolytic cleavage of a loosely bound H
2
O
2
, resulting in the formation of 
compound I and H
2
O. Compound I is not stable and is proposed to react with the halide to form a 
hypothetical ferric hypohalite adduct, termed compound X. Decomposition of compound X is 
proposed to produce a halonium ion (X
+
) along with ferric enzyme. In aqueous solution, the 
hypohalous acid would be expected to form from X
+
. Recently, based on substrate-bound 
structures, a modified mechanism for compound I formation has been proposed [47]. Instead of 
direct formation of compound I, a ferric-hydroperoxyl species (compound 0) is proposed to be 
formed first; the Glu 183 then protonates the distal oxygen of compound 0, followed by the 
formation of compound I along with cleavage of the O-O bond and release of one water 
molecule (Fig. 1.11). 
 
 22 
 
Figure 1.10: Catalytic cycle of haloperoxidases.  
 23 
 
Figure 1.11: Proposed mechanism of compound I formation by chloroperoxidase.  
 
 
 
 24 
1.3.5 The peroxidase-cyclooxygenase superfamily: 
The peroxidase-cyclooxygenase superfamily represents one of the two main groups of 
heme-dependent peroxidases in biology. Peroxidase-cyclooxygenase and peroxidase-catalase 
superfamilies arose independently; therefore, the structures of both the protein and prosthetic 
group are distinct between these two groups [18]. The peroxidase-cyclooxygenase superfamily 
was originally called the mammalian peroxidases, and was later changed to the animal 
peroxidase family. However, in recent years, even plant, fungal and bacterial representatives 
were found and assigned to this superfamily necessitating yet another change in nomenclature. 
Though this superfamily has been subdivided into seven clades, it is still identified by its most 
famous members (myeloperoxidase [MPO], eosinophil peroxidase [EPO], lactoperoxidase [LPO] 
and thyroid peroxidase [TPO]) all of which are found in animals. Because this superfamily is not 
phylogenetically related to the superfamily containing KatG, this section will only consider the 
intensively investigated representative, myeloperoxidase (MPO). An interesting feature of many 
representatives from the peroxidase-cyclooxygenase superfamily is novel covalent linkage 
between the porphyrin cofactor and protein.  In this, MPO is an excellent example. 
 
1.3.5.1 Structure of MPO: 
Myeloperoxidase is abundant in neutrophils but also is expressed in monocytes and certain 
types of macrophages, where it participates in innate immune defense mechanisms by generating 
reactive oxidants and diffusible radical species [49]. The mature enzyme exists as a homodimer, 
where each subunit is dominated by ?-helices with few ?-strands.  The subunit is composed of 
two polypeptide chains resulting from the post-translational excision of a peptide fragment from 
a single polypeptide precursor [50, 51]. The catalytically essential residues Gln 91, His 95 and 
 25 
Arg 239 occupy the distal heme pocket; in the proximal site, the heme iron is coordinated 
through the His 336 imidazole resiude.  This ligand forms an H-bond with the amide carbonyl 
oxygen of Asn 421. One Ca
2+
 ion has been identified and its coordination is essential for 
maintaining the distal architecture. 
 
A common, but not universal, feature for members of this superfamily is the covalent 
association of the heme with the protein itself: pyrrole rings A and C of the heme group are 
methylated, allowing formation of ester linkages with the carboxyl groups of Glu 242 and Asp 
94. Additionally, the ?-carbon of the vinyl group on pyrrole ring A forms a covalent bond with 
the sulfur atom of Met 243 (Fig. 1.12). As one might anticipate, this produces unique spectral 
properties for the heme. In the ferric state, the Soret band shows an absorption maximum around 
428 nm and relatively strong absorption bands in the visible region are also detected, which are 
responsible for the characteristic green color of the enzyme [52]. This unique post-translational 
modification gives MPO peculiar redox properties, allowing oxidize Cl
- 
[53]. 
 
1.3.5.2 Enzymatic properties and catalytic mechanism: 
Under physiological conditions, MPO catalyzes the oxidation of halides (X
-
) to the 
corresponding hypohalite anions (OX
-
). To do this the ferric enzyme is oxidized by H
2
O
2
 to form 
compound I. Distal His 95 and Arg 239 are proposed to orient and bind the H
2
O
2
, allowing the 
heterolytic fission of the O-O bond much as has been observed for other peroxidatic enzymes. In 
addition to its halogenation activity, MPO is also able to catalyze more typical oxidations of 
other peroxidatic electron donors. Therefore, once compound I is formed, MPO can catalyze 
 
 26 
 
 
 
Figure 1.12: Structure of human MPO. 
Overall structure (A), and active site (B) of human MPO. Different colors highlight the structures 
from two different polypeptides (PDB:1CXP) [54].  
 
 27 
 
 
 
Figure 1.13: Catalytic cycles of MPO.  
Halogenation cycles: reaction 1 and 2; Peroxidase cycles: reaction 1, 3 and 4 [49]. 
 28 
 either halogenation or canonical peroxidation (Fig. 1.13). In the halogenation cycle, compound I 
is reduced back to the ferric state by the oxidation of halide (X
-
) substrates. In the peroxidase 
cycle, compound I is reduced back to the native state in two sequential one-electron steps using 
another exogenous reducing substrate (AH
2
). Presumably, the operating cycle is determined by 
the availability of substrate. 
 
1.3.6 The peroxidase-catalase superfamily: 
The first systematic classification of the enzymes of this superfamily was published in 
1992 [38] and divided its members into three classes. In contrast to other peroxidase 
superfamilies, this subdivision has remained unchanged since that time. Class I includes 
cytochrome c peroxidases, ascorbate peroxidases and catalase-peroxidases (KatG). Class II 
contains fungal peroxidases, predominantly lignin and manganese peroxidases. Class III includes 
the secretory plant peroxidases. Though the subdivisions have not changed, the name of the 
overall superfamily was recently changed from the superfamily of plant, fungal, and bacterial 
heme peroxidases to peroxidase-catalase superfamily in order to reflect the main enzymatic 
activities of its members [21]. 
 
The Class I members are widely distributed in nature with the exception of animals. 
Though the physiological roles of the peroxidase-catalases vary considerably, the class I 
enzymes serve exclusively as H
2
O
2 
scavengers [55-57]. The class II peroxidases are found 
exclusively in fungi and have a major role in degradation of lignin [58, 59]. Class III are mostly 
plant-secreted glycoproteins with roles in a variety of physiological processes of the plant life 
 29 
cycle such as cell wall metabolism [60], wound healing [61, 62], auxin catabolism [63], H
2
O
2
 
scavenging, and defense against pathogens [64].  
 
1.3.6.1 Class III: Plant secretory peroxidases: 
The most intensively investigated representative of this family is horseradish peroxidase  
(HRP) from Armoracia rusticana. In this section, the discussion of structure and catalytic 
mechanism of class III peroxidases will focus on HRP. 
 
1.3.6.1.1 Structure of HRP: 
Crystal structures reveal that HRP has ten helices assigned A-J which produce a fold that is 
highly conserved in the Class III family, and several unique features: a long insertion between 
helices F and G; a short antiparallel ??sheet preceding the C177-C209 disulfide bridge; and a 
three-residue insertion between the H and I helix [65]. In HRP, all the N-glycosylation sites 
positioned in loop regions are pointing away from the molecule except Asn 268 which is located 
right at the end of the I helix, indicating the purpose of glycosylation is to increase solubility of 
the enzyme [65]. Recent reports indicate that glycan may stabilize the protein and adjust the 
substrate access [66-68]. In addition, there are four disulfide bonds between cysteine residues 11-
19, 44-49, 97-301 and 177-209. Of those, the disulfide bond in the distal heme-binding domain is 
proposed to dedicate to stabilizing active site structure [64]. 
 
In the active site, the ferric heme b prosthetic group is ligated to the protein though His 
170, which also hydrogen bonds to the Asp 247 carboxylate group. Three strictly conserved 
residues are located at the distal site: His 42 serves as a general base/acid; Arg 38 is essential for 
 30 
the formation and stabilization of compound I; Phe 41 prevents substrate access to the ferryl 
oxygen of compound I [69]. In addition, Asn 70 is strongly conserved in the peroxidase-catalase 
superfamily and has been proposed to maintain the basicity of the distal His (His 42) [70].  In 
both the distal and proximal sites, two hepta-coordinate Ca
2+
 ions are required for both structural 
stability and catalytic functions [69, 71, 72] (Fig. 1.14). Loss of Ca
2+ 
or disruption of its 
interactions results in the B-helix bearing distal His shift, allowing the distal His to directly 
coordinate with the heme iron. Therefore, the Ca
2+
-deficient HRP is inactive, but this is 
reversible and can be remedied by addition of Ca
2+
 [69, 71, 72]. 
 
1.3.6.1.2 Catalytic mechanism of HRP: 
The mechanism of catalysis of HRP has been extensively investigated [73, 74]. Similar to 
those of other peroxidases we described previously, the first step in the catalytic cycle is 
generation of compound I. Once compound I is formed, typically aromatic electron donors 
reduce compound I to compound II, then compound II is reduced back to the resting state of the 
enzyme by another equivalence of reducing substrate. Although the formation of these 
intermediates has been proposed for a number of catalytic heme enzymes, these species are 
formed only transiently and have been somewhat elusive to identify until recent years. The 
structures of the HRP catalytic cycle intermediates have been obtained at high resolution by X-
ray crystallography, allowing direct and accurate comparison of the bonding interactions in the 
different intermediates [75]. The catalytic cycle of HRP with the structures of the intermediates 
is summarized in Fig. 1.15.   
 
 
 31 
 
 
 
Figure 1.14: Structure of HRP. 
Overall structure of horseradish peroxidase (HRP) (PDB: 1h5a) (A) and the active site of HRP 
(B) [75]. 
 
 32 
 
 
 
Figure 1.15: Peroxidase cycle of HRP. 
 
 
 
 33 
1.3.6.2 Class II: Manganese, lignin and versatile peroxidases: 
Extracellular Class II heme peroxidases are currently known only in fungi and have their 
principle function in lignin degradation [76]. There are three main groups secreted by 
Basidiomycetes or so-called white rot fungi: manganese peroxidases (MnP), lignin peroxidases 
(LiP), and versatile peroxidases (VP).  
 
1.3.6.2.1 Structure and function of Class II peroxidases: 
The degradation of lignin to CO
2
 by this small and specialized group of fungi is a major 
route for the recycling of photosynthetically fixed carbon [77]. Lignin is chemically recalcitrant 
to catabolism by most organisms because of its complex, heterogeneous and irregular structure. 
In nature, only the Basidiomycota are able to degrade lignin efficiently. They produce four major 
groups of enzymes for lignin degradation: LiP, MnP, VP that belong to Class II peroxidase 
family and a laccase, which is a multi-copper-containing protein. 
 
Along with the two other peroxidases, LiP catalyzes the H
2
O
2
-dependent oxidative 
degradation of lignin, but the mechanism by which this occurs is poorly defined [78, 79].  In 
addition, LiP displays classical peroxidatic catalytic promiscuity with respect to the electron 
donor substrate.  The enzyme has been shown to oxidize phenolic compounds (especially 
veratryl alcohol) as well as a broad range of nonphenolic aromatic compounds. Interestingly the 
reach of LiP in terms of the redox potential of compounds it will oxidize extends as high as 1.4 V 
[80]. The pH optimum (~3) for the enzyme is strikingly low considering the structure of its 
active site. Crystal structures for LiP reveal eight major and eight minor ?-helices with a few ? 
structures in the proximal domain [81]. The overall folding and active site environment of LiP 
 34 
are highly similar to those of HRP (Fig. 1.16A). In addition, LiP contains three glycosylation 
sites, which have been indicated to protect the C-terminal peptide from proteolysis [77]. Similar 
to HRP, LiP has four disulfide bridges and two structural calcium ions which are essential for 
maintaining the structure of the active site [81]. The location of the two Ca
2+
 coordination sites 
are similar to those observed in HRP; however, the disulfide bond that enhances the stability of 
the distal Ca
2+
 binding site in HRP is absent in LiP, making the coordination state and enzyme 
activity much more sensitive to the loss of the Ca
2+ 
ion in LiP.  
 
Interestingly, a novel post-translational hydroxylation of Trp 171 (C?) has been identified 
[81]. This is proposed to have an important role in the binding and oxidation of aromatic 
substrates through long-range electron transfer [82-84]. This Trp residue and the electron-
deficient heme are purported to account for the high reduction potentials of the enzyme?s ferryl 
intermediates [85]. These endow LiP with the unique ability to catalyze oxidative cleavage of C-
C bond and ether bonds in non-phenolic aromatic substrates. 
 
Manganese peroxidase is an acidic glycoprotein. The overall structure of MnP is similar to 
that of LiP [86]. Two heptacoordinated structural calciums with similar positions and functions 
as those observed in LiP [87] (Fig. 1.16B). The active sites of MnP and LiP are virtually 
identical to one another. MnP does possess an additional disulfide bond in addition to those four 
that are also seen in LiP.  This unique fifth disulfide is proposed to assist in Mn
2+
 binding by 
pushing a C-terminal peptide segment away from the main body of the protein [77]. A bound 
Mn(II) ion has been identified at the surface of the protein coordinated to the carboxylates of Glu 
35, Glu 39 and Asp 179, one of the heme propionates and two water molecules.  As described in 
 35 
greater detail below, MnP utilizes Mn
2+
 as a direct substrate, the oxidation product of which 
mediates the oxidation of secondary substutrates. 
 
Structurally, versatile peroxidases (VP) resemble LiP more closely than MnP. The 
available structure shows a 12 helices, 4 disulfide bonds, a heme pocket containing the 
characteristic proximal and distal triads, 2 structural Ca
2+
 sites, and a Mn (II) binding site (Fig. 
1.16C) [88]. The solvent accessible Trp 64 is essential in the oxidation of high redox potential 
aromatic substrates via a long-range electron transfer pathway similar to that observed for LiP 
[89]. 
 
1.3.6.2.2 Catalytic cycles of Class II peroxidases:  
LiP has a typical peroxidase catalytic cycle similar to that of horseradish peroxidase (Fig. 
1.15). The primary different is in the involvement of long-range electron transfers for LiP. In a 
sense, MnP utilizes a similar long-range electron transfer strategy as that observed for LiP; 
however, instead of oxidizing a surface tryptophanyl residue, MnP catalyzes the oxidation of 
Mn(II) to Mn(III) which in turn may be used to oxidize numerous monomeric phenols including 
dyes as well as phenolic lignin model compounds [90]. The overall characteristics of the cycles 
of LiP, VP, and MnP are highly similar to those of HRP (Fig. 1.17). Another difference between 
LiP and MnP is oxidation of non-phenolic substrates by MnP involves the formation of reactive 
radicals in the presence of second mediators (e.g., thiols) [91]. The mechanism employed by VP 
is essentially the same as that of LiP except that VP can also utilize Mn(II) as a mediatory 
substrate much as MnP does.   
 
 36 
 
 
 
Figure 1.16: Structure of Class II peroxidases. 
Overall structures of LiP (PDB: 1LGA)(A) [81], MnP (PDB: 1MNP) (B) [86] and VP (PDB: 
2BOQ) (C) [88], and the active sites of LiP (D), MnP (E) and VP (F).
 37 
 
 
Figure 1.17: Catalytic cycle of manganese peroxidase (MnP).  
 
 
 
 
 
 
 
 
 
 
 
 38 
 
1.3.6.3 Class I peroxidases: 
1.3.6.3.1 Introduction of Class I peroxidases: 
Class I peroxidases include intracellular prokaryotic and eukaryotic non-glycosylated heme 
peroxidases without Ca
2+
 [38, 92]. Extensive gene sequence analysis of all three classes indicate 
that the whole peroxidase-catalase superfamily originated from ancestral Negibacteria, and 
during evolution the already diversified peroxidase genes were transferred via two ancestral 
endosymbiotic events to newly formed eukaryotic cells [19, 93]. Class I segregated at an early 
stage from a common ancestor of the class II and class III peroxidases [94]. However, the 
evolutionary lineage of CCP and APX from the two-domain KatG remains contested [38, 92].  
 
Ascorbate peroxidases are predominantly observed among green algae and higher plants. 
Their primary physiological role is to scavenge H
2
O
2 
as part of the ascorbate-glutathione cycle 
[55, 95]. Ascorbate peroxidase plays an important role in root nodules of leguminous plants and 
chloroplasts in general. Given that O
2
 is a known inhibitor of photosynthesis and can deactivate 
nitrogenase, a peroxide degradation mechanism that avoids generating O
2
 is particularly 
advantageous.   
 
The subfamily of CCPs can be divided into two groups: mitochondrial and non-
mitochondrial. The enzyme from Saccharomyces cerevisiae is located in the intermembrane 
space of mitochondria where it decomposes H
2
O
2
 generated by the mitochondrial respiratory 
chain [96]. In addition to acting as a H
2
O
2
 scavenger, CCP appears to be integral to the oxidative 
stress response to methanol utilization in the methylothropic yeast [97]. The enzyme from 
 39 
pathogenic basidiomycetes has been implicated in the bacterium?s resistance against exogenous 
oxidative stress [98]. 
 
1.3.6.3.2 Structure and Catalytic function of APX and CCP: 
As with all members of the peroxidase-catalase superfamily, the structures of all class I 
enzymes are dominated by the ??helix. As a percentage, KatGs exhibit the lowest contribution 
from ??helix, whereas the APXs have the highest with CCPs falling in between. All class I 
enzymes carry heme b as a cofactor, and in each case, it is non-covalently held by the protein 
that carries it. The distal Arg-Trp-His and proximal His-Asp-Trp triads of the active sites are 
strictly conserved among class I enzymes not only in sequence, but also in their 
position/orientations within the active site (Fig. 1.18) [2]. There is some evidence that the distal 
Arg may occupy two positions: one ?out? and one ?in? [99], where only Arg occupying the ?in? 
position can form H-bond with the iron-linked ferryl oxygen atom. As elsewhere this interaction 
is proposed to stabilize the ferryl center of compound I [99].  Crystal structures as well as 
spectroscopic properties of the enzymes indicate several differences between CCP and other 
peroxidases. The typical compound I (ferryl porphyrin radical) does not accumulate in CCP due 
to facile electron transfer from an active site, forming instead a ferryl tryptophanyl radical 
intermediate referred to by some as ?compound ES? [100]. In addition, two fully conserved 
methionines (230 and 231) serve to stabilize the Trp radical [101]. In contrast, a typical 
compound I (ferryl porphyrin radical) intermediate is observed upon reaction of APX with H
2
O
2
.  
In this protein, a potassium ion bound adjacent to the proximal Trp not only stabilizes the 
conformation of the APX active site [102], but also disfavors Trp oxidation to its corresponding 
radical intermediate. Beyond this, the catalytic cycles APX and CCP are highly similar to those 
 40 
observed for class II and III peroxidases (Fig. 1.15 and 1.17).  One notable feature in both cases 
is the identity of the preferred electron donor substrate.  Reduced cytochrome c and ascorbate are 
both strikingly poor electron donors for all other peroxidases.  In these instances, CCP and APX 
have clearly adapted to employ the electron donor that is available in the cellular environment of 
each one. 
 
1.4 Catalase-peroxidases (KatGs): 
Phylogenetic analyses suggest the duplication event evident in the KatG gene may have 
occurred in the later phases of evolution, where the individual and distinguishing features of the 
peroxidase families were already formed [102]. The KatGs are broadly distributed among 
archaea, bacteria and lower eukaryotes.  It appears that their primary purpose across these 
organisms is to degrade H
2
O
2
. Despite all their similarities to other peroxidase-catalase 
superfamily members (especially class I enzymes), KatGs are unique in that they use single 
active site to catalyze two fundamentally different reactions. 
 
1.4.1 The biomedical benefits of understanding KatG structure and function: 
The vast majority of superfamily members are monomeric, and with the lone exception of 
KatG, the subunits of these enzymes are composed of a single domain.  As a consequence, the 
novel two-domain composition of KatGs provide a rare opportunity to evaluate mechanisms in 
which structures peripheral to and distant from an active site influence its architecture and 
catalytic ability. This aspect of the structure/function relationship is rarely studied and poorly 
understood. 
 
 41 
 
 
Figure 1.18: Structure of Class I peroxidase. 
Overall structure of CCP (PDB: 2CYP)(A), APX (PDB: 1OAF) (B) and the active site of CCP 
(C), APX (D) [103, 104]. 
 
 
 42 
There are substantial biomedical benefits to be derived from an understanding of KatG 
structure and function.  First, KatG from Mycobacterium tuberculosis has a central role in the 
activation of isoniazid (INH), a front-line antitubercular agent. In M. tuberculosis, INH is 
activated by KatG, leading to the formation of an INH-NAD adduct [105-108]. This adduct 
inhibits InhA, the enoyl-ACP reductase of the fatty acid synthase type II system, which is the 
central enzyme complex in mycolic acid biosynthesis [109, 110]. Consequently, inhibition of 
mycolic acid production interferes with mycobacterial cell wall construction and ultimately 
results in cell death [109, 110]. It has been estimated that 70% of isoniazid resistant strains carry 
mutations to the gene for KatG. The mechanisms by which many of the commonly observed 
mutations compromise isoniazid activation are not known.  Clearly, a better understanding of 
KatG structure and function will be beneficial for resolving these questions. 
 
In addition, KatGs have been implicated as virulence factors in organisms like Yersinia 
pestis [111-113], Legionella pneumophila [114], and enterohemorrhagic E. coli [115, 116]. 
These organisms produce a periplasmic or extracellular form of KatG. Because a central strategy 
of antibacterial responses of plants, mammals and other higher eukaryotes relies on the 
production of copious quantities of reactive oxygen species, particularly H
2
O
2
, the advantage to 
be gained from a robust extracellular catalase is considerable.  The oxidative burst of neutrophils 
and macrophages rapidly produces high levels of O
2
?? which disproportionate to H
2
O
2
. In 
neutrophils, H
2
O
2
 is used by myeloperoxidase to oxidize chloride to hypochlorous acid.  In 
macrophages, the rapid generation of ?NO along with O
2
?? produces peroxynitrite anion. The 
ability to rapidly degrade H
2
O
2
 and other peroxides like peroxynitrite is essential for pathogenic 
bacteria employ as a defense against the host?s immune response. Therefore, the periplasmic 
 43 
location of these enzymes is an essential strategy that allows access to the target substrate before 
it has had an opportunity to damage essential cellular systems.  The connection between KatG 
structure and its utility in this role remains to be explored, and a greater knowledge of KatG 
structure will be necessary in order to exploit this enzyme as a target to hobble bacterial 
antioxidant defenses, making pathogens like E. coli O157:H7 or Yersinia pestis more vulnerable 
to the hosts? own natural antibacterial defenses. 
 
1.4.2 Structure of KatGs: 
The catalase-peroxidases exist as either homodimers or homotetramers [117]. Each subunit 
is composed of two structurally homologous domains [118]. Neither domain bears any 
resemblance to monofunctional catalases; rather, both domains strongly resemble 
monofunctional peroxidases, bearing the same topological structure and similar helical structure 
around the active sites [119-122]. Structural data and amino acid sequence alignments indicate 
that there are 10 major helices in each domain, similar to other class I members. Interestingly, 
the heme-binding consensus sequence is only observed in the N-terminal domain, and hence, this 
domain contains the only functional active site. In contrast, the C-terminal domain lacks the 
heme-binding motif, and as a result, does not bind heme or catalyze a reaction.  
 
At the active site level, the available crystal structures reveal that the KatGs have the 
proximal and distal conserved amino acids at almost identical positions as in other class I 
peroxidases. In particular, both the triads His/Trp/Asp and His/Arg/Trp are strongly conserved 
and almost in superimposable positions found in other class I peroxidases (Fig. 1.19) [119-122]. 
In addition, a conserved Asp 135 participates in the extensive KatG-typical hydrogen-bond  
 44 
 
 
Figure 1.19: Comparisons of the overall structures and the active sites of monofunctional 
peroxidases, KatG and monofunctional catalases. 
APX, (PDB: 1OAF) (A and D, respectively), catalase-peroxidase (PDB: 1SJ2)  (B and F, 
respectively) and HPII from E.coli, (PDB: 1IPH) (C and G, respectively). 
 
 
 
 
 
 
 
 
 45 
network at the distal heme pocket. Asn 136 is conserved in all members of the peroxidase-
catalase superfamily and supports this network through a hydrogen bond to the distal His. On the 
basis of site-directed mutagenesis and spectroscopic studies, it has been suggested that the 
catalase-unique reaction requires this precisely aligned H-bond network [123-125]. Another 
obvious difference between KatG and monofunctional peroxidase is that there is a pronounced 
funnel channel to the active site in KatG that is reminiscent of monofunctional catalases. In 
KatG, the acidic residues are critical for stabilizing the solute matrix and orienting the water 
dipoles in the channel. Accordingly, exchange of the acidic residues affects catalase but not 
peroxidase activity [126]. 
 
Most KatGs do not contain any disulfide bonds.  Extracellular KatGs secreted by certain 
fungi stand as an exception. Indeed, the recently solved structure for a eukaryotic KatG from 
Magnaporthe grisea shows two disulfide bridges between the N-terminal domains of two 
subunits [127]. These cysteines, and presumably the disulfide bonds which result, appear to be 
fully conserved but only in the group of the secreted fungal KatGs (Fig. 1.20). They are also 
absent in the other members of class I peroxidases such as CCP and APX. These disulfide bonds 
significantly increase the structural integrity of the enzyme, making the N-terminal domain more 
stable than the C-terminal domain [127]. 
 
Although KatG belongs to the class I peroxidase family, none of the other class I members 
(e.g., CCP and APX) has appreciable catalase activity. Given the high similarity of the 
monofunctional peroxidase and KatG active sites, it is reasonable to suggest that KatG-unique 
protein structures external to the active site have an important role in fine-tuning the active site  
 46 
 
 
Figure 1.20: Two novel disulfide bridges between the N-terminal domains of two subunits 
in KatG. 
KatG from M. Grisea KatG 2 (PDB: 3UT2) shown in the overall structure (A) and zoom-in view 
(B) [127]. 
 47 
for its bifunctional abilities. In support of this hypothesis, sequence alignments [118, 128] and 
the four available crystal structures of the KatGs [119-122] reveal several structural features that 
are unique to these enzymes. Two of them are Large Loops 1 and 2, LL1 and LL2, respectively 
(Fig. 1.21). Relative to the other class I enzymes, each one adds about 35 amino acids to the 
interhelical loops connecting the D and E helices and F and G helices, respectively.  Evaluation 
of a wide array of substitution and deletion variants targeting the LL2 structure suggests that this 
loop structure may support hydrogen-bonded networks critical for reactions involving H
2
O
2
 and 
may regulate access of electron donors to the active site [129, 130].  
 
A striking feature of LL1 is that it contains the invariant Tyr (Y226 by E. coli KatG 
numbering), which participates in a novel Met-Tyr-Trp covalent adduct located on the distal side 
of the heme active site. This adduct has been observed in the crystal structures [119-121, 131] 
(Fig. 1.22) as well as by mass spectrometric peptide mapping [132-135]. Elimination of LL1 or 
substitution of Y226 eliminates catalase activity, but deletion variants are substantially more 
active as peroxidases, up to an order of magnitude increase to the rate in same cases [128, 136]. 
In addition to orienting this invariant tyrosine for participation in the essential covalent adduct, 
LL1 also appears to regulate the access of reducing substrates to the heme edge [128, 136].  
 
1.4.3 Catalytic cycles of KatGs: 
KatGs are capable of H
2
O
2
 decomposition by catalatic and peroxidatic mechanisms, and 
they do so using a single active site. In the classical model, both catalase and peroxidase cycle 
start with heterolytic reduction of H
2
O
2
 along with oxidation of ferric enzyme to compound I 
(Fig. 1.23A). Once compound I is formed, it can be reduced to the resting state either 
 48 
 
 
 
Figure 1.21: Structure of catalase-peroxidase. 
KatG from M. tuberculosis (PDB: 1SJ2) [131]. 
 
 
 
 
 49 
 
 
 
 
 
Figure 1.22: The covalent adduct in KatG. 
KatG from M. tuberculosis (PDB: 1SJ2) (Numbering reflects E.coli structure of KatG) [131]. 
 
 
 
 
 50 
 
 
Figure 1.23: Classical model of catalytic cycle of KatG (A), modified model of catalase 
mechanism of KatG (B). 
 
 
 
 
 
 51 
through a catalase or peroxidase cycle. However, this model cannot explain the role of M-Y-W 
adduct in the catalytic mechanism. In addition, recently, the Goodwin laboratory has reported the 
stimulation of catalase activity by classical peroxidatic electron donors rather than inhibition, 
particularly under conditions favorable to peroxidase activity. This is wholly inconsistent with 
the classical paradigm. One mechanism proposed for the catalatic turnover of KatG that accounts 
at least for the former observation has been proposed. By this mechanism compound I is 
converted to compound I* wherein the unique covalent adduct serves as an endogenous electron 
donor. An additional equivalent of H
2
O
2
 then converts compound I* to compound III*. Finally, 
the adduct radical serves as an electron acceptor returning the enzyme to its ferric state with the 
concomitant release of O
2
 [137-140] (Fig. 1.23B). 
 
1.4.4: The roles of C-terminal domain: 
As mentioned above, all other members of the peroxidase-catalase superfamily are single 
domain proteins.  Thus, the C-terminal domain, a structure ? 30? from the active site, is unique 
to KatG. The structure is conserved across all KatGs.  Based on its sequential and structural 
similarity to the N-terminal domain, the C-terminal domain is proposed to have originated from a 
gene-duplication and fusion event [118, 128, 141]. However, its roles in catalysis are not 
immediately obvious. Nevertheless, the Goodwin laboratory has observed that KatG without its 
C-terminal domain (KatG
N
) is no longer expressed in a soluble form but is instead found in 
inclusion bodies. In addition, following denatured purification, refolding and reconstitution, 
KatG
N
 shows a shifted active site structure, allowing what appears to be the direct coordination 
of the heme iron by the distal histidine (H106 by Escherichia coli numbering) [142]. 
Consequently, KatG
N
 has neither catalase nor peroxidase activity. It has been suggested that the 
 52 
C-terminal domain might serve as a platform to direct the proper folding of the N-terminal 
domain and facilitate dimerization [143]. This is supported by previous observations from the 
Goodwin laboratory [142, 144]. Introducing the separately expressed and isolated C-terminal 
domain (KatG
C
) results in restructuring KatG
N
 that is evident from a return of high-spin 
coordination states typical of the wild-type enzyme as well as the restoration of catalase and 
peroxidase activities [144]. Therefore, the C-terminal domain is an integral component of KatG 
intimately tied to its structure and function; this is also evident from a recent thermodynamic 
study of KatGs showing that the N- and C-terminal domains unfold as a single unit rather than 
independently [145].  
 
However, the mechanism by which the C-terminal domain modulates the active site 
conformation is still not clear. Inspection of the crystal structures for various KatGs reveals that 
there are clear interactions between the two domains, which are proposed to contribute to the 
stability of the overall quaternary structure of the enzyme [119, 120, 122, 146]. As an extension 
of this hypothesis, it is reasonable to propose that the C-terminal domain, through these 
interactions, directs conformational adjustments in the adjacent domain including the active site 
itself. In order to evaluate the individual structures and mechanisms employed by the C-terminal 
domain, we focused our attention on a conspicuous structural element of this domain, the I?-helix 
(Chapter 2). Our results show that this helix is a central component of a platform to facilitate 
folding events and/or conformational adjustment in the N-terminal domain. Upon establishing 
the critical contributions of this structure, we turned our attention to understanding the 
mechanism(s) by which this remote structure communicates with and controls the active site 
(Chapter 3). The C-terminal domain has a unique evolutionary history.  It was born in a gene-
 53 
duplication and fusion event.  Either because the maintenance of its own catalytic function was 
unnecessary, or because a catalytically active C-terminal domain was a hindrance to the overall 
function of KatG, the specific structures necessary for heme binding and catalytic transformation 
of H
2
O
2
 were eliminated.  Nevertheless, the general scaffold for a heme-based catalyst was 
retained to near perfection.  We surmised that this provided an ideal avenue toward the creation 
of new heme protein-based catalysts.  Our efforts and accomplishments in this direction are 
given in Chapter 4. Finally, our previous observations have shown that the C-terminal domain is 
essential for both catalase and peroxidase activities of KatG. It is also known that a novel 
covalent adduct in the N-terminal domain is critical, but only to the catalase activity of the 
enzyme. More importantly, the establishment of this adduct is both heme and peroxide 
dependent.  This would suggest that the C-terminal domain is also essential for formation of the 
covalent adduct, but this has never been investigated.  By way of a novel procedure to evaluate 
covalent adduct formation, we have determined that establishment of the covalent adduct cannot 
occur without the C-terminal domain (Chapter 5). It is noteworthy that although the C-terminal 
domain accounts for nearly half the structure of this important enzyme, prior to these 
investigations it remained the least understood of all of KatG?s novel structural features. 
Together, these studies represent a substantial advance in understanding the function and 
potential applications of this important structure, and we anticipate that they will contribute not 
only to answering fundamental questions surrounding protein structure and function, but also to 
providing insight into the specific biomedical importance of KatG, in particular, its central 
involvement in Mycobacterial resistance to antibiotics and the virulence of pathogenic bacteria. 
 
 
 54 
 
 
 
 
 
CHAPTER TWO 
 
 
INTEGRAL ROLE OF THE I?-HELIX IN THE FUNCTION OF THE ?INACTIVE? C-
TERMINAL DOMAIN OF KATG 
 
2.1 Introduction: 
Catalase-peroxidases (KatGs) have been the subject of considerable attention since the 
discovery of their role in the activation of the frontline anti-tubercular agent isoniazid in 
Mycobacterium tuberculosis [105-107]. Catalase-peroxidases have also been implicated as 
virulence factors for some highly virulent pathogens such as Yersinia pestis [111-113], 
Legionella pneumophila [114]. Nevertheless, the mechanism by which these enzymes may serve 
as virulence factors have not been illuminated. Clearly, there are important benefits to be derived 
from understanding the structure and function of KatG. 
 
Catalase-peroxidases are distributed among archaea, eubacteria, fungi and protists [93] 
primarily as hydrogen peroxide scavengers [147]. These enzymes use a single active site to 
degrade H
2
O
2
 by either a catalase or peroxidase mechanism. KatGs are either homodimers or 
homotetramers [148], and each subunit has two structurally homologous domains [118]. Neither 
domain bears any resemblance to monofunctional catalases; rather, both domains strongly 
resemble monofunctional peroxidases, bearing the same topological structure and similar helical 
structure around active sites [119-122]. Indeed, catalase-peroxidases belong to class I of the 
 55 
peroxidase-catalase superfamily [118]. However, none of the other class I members (e.g., 
cytochrome c peroxidase [CCP] and ascorbate peroxidase [APX]) has appreciable catalase 
activity. Sequence alignments [118, 128] and the four available crystal structures of the KatGs 
[119-122] reveal several structural features that are unique to catalases-peroxidases. Two of them 
are Large Loop 1 (LL1) and Large Loop 2 (LL2). Relative to the class I enzymes, each one adds 
about 35 amino acids to the interhelical loops connecting the D and E helices and F and G 
helices, respectively. Likewise, other peroxidases are single-domain proteins, making the 300-
amino acid C-terminal domain another unique feature of KatGs. It may be the persistence of the 
longer interhelical loops, both of which are essential for catalase but not peroxidase activity [132, 
134, 149-151], along with the necessity for a precisely ordered active site to support catalase 
activity [145] that has compelled the continued presence of the C-terminal domain in KatG but 
not its most closely related neighbors CcP and APx. 
 
The C-terminal domain is ? 30? from the active site. It is conserved across all catalase-
peroxidases, and based on its sequential and structural similarity to the N-terminal domain, is 
proposed to have originated from a gene-duplication and fusion event [118, 128]. However, its 
roles in catalysis are not immediately obvious. It bears a striking resemblance to monofunctional 
peroxidases [119-122, 143], but it neither binds heme nor catalyzes any discernable reaction. It is 
proposed that the class I enzymes were all derived from a two-domain predecessor, suggesting 
that deletion of one of the domains was a necessary step in the development of CcP and APx 
[92]. In this, the C-terminal domain has the appearance of a vestigial structure. 
 
 56 
Contrary to this notion, KatG without its C-terminal domain (KatG
N
) is no longer 
expressed in a soluble form but is instead found in inclusion bodies. In addition, following 
denatured purification, refolding and reconstitution, KatG
N
 shows a shifted active site structure, 
allowing what appears to be the direct coordination of the heme iron by the distal histidine (H106 
by Escherichia coli numbering) [142]. Consequently, KatG
N
 has neither catalase nor peroxidase 
activity. It has been suggested that the C-terminal domain might serve as a platform to direct 
proper folding of the N-terminal domain and facilitate dimerization [143], which is supported by 
our previous observation. Introducing the separately expressed and isolated C-terminal domain 
(KatG
C
) results in restructuring of KatG
N
 that is evident from a return of high-spin coordination 
states typical of wild-type enzyme as well as the restoration of catalase and peroxidase activities 
[144]. That the C-terminal domain is an integral component of KatG intimately tied to its 
structure and function is also evident from a recent thermodynamic study of KatGs showing that 
the N- and C-terminal domains unfold as a single unit rather than independently [145]. 
 
The I?-helix in the C-terminal domain is highly conserved and appears at the interface 
between the two domains. It makes contacts with several elements from the N-terminal domain 
of the adjacent subunit using strongly conserved Leu 690, Arg 691, Glu 695, and Tyr 697. As 
such, it has all the appearances of a platform by which the C-terminal domain may direct 
structural adjustments to the N-terminal domain thereby modulating active site conformation. To 
evaluate this hypothesis, we produced variants substituting alanine for these conserved residues. 
Single, double, and quadruple alanine substitution variants of intact KatG were expressed, 
isolated and characterized, and the effects of these substitutions on active site structure, stability 
and catalytic bifunctionality were evaluated. In addition, the analogous variants were also 
 57 
expressed and isolated for the stand-alone C-terminal domain (KatG
C
). The ability of these 
KatG
C
 variants to restore KatG
N
 to an active conformation was evaluated. Results from both sets 
of variants demonstrate that the I?-helix is central to the ability of the C-terminal domain to direct 
structural adjustments in the adjacent N-terminal domain. This supports the hypothesis that the 
C-terminal domain serves as a platform to direct N-terminal domain active site conformation and 
bifunctionality. 
 
2.2 Materials and methods: 
2.2.1 Materials: 
Hydrogen peroxide (30%), imidazole, hemin, ampicillin, chloramphenicol, sodium 
dithionite, phenylmethylsulfonyl fluoride (PMSF), 2,2?-azino-bis(3-ethylbenzthiazoline-6-
sulfonate) (ABTS), and guanidine hydrochloride (GuHCl) were purchased from Sigma (St. 
Louis, MO). Isopropyl-?-d-thiogalactopyranoside (IPTG), urea, mono- and dibasic sodium 
phosphate, acetic acid, and sodium acetate were obtained from Fisher (Pittsburgh, PA). 
Bugbuster and benzonase were purchased from Novagen (Madison, WI). All restriction enzymes 
were purchased from New England Biolabs (Beverly, MA). All oligonucleotide primers were 
purchased from Invitrogen (Carlsbad, CA). All E. coli strains (BL-21-Gold [DE3] pLysS and 
XL-1 Blue), Pfu polymerase, and T4 DNA ligase were obtained from Agilent (La Jolla, CA). 
Nickel-nitrilotriacetic acid (Ni-NTA) resin was purchased from Qiagen (Valencia, CA). 
Desalting 10DG chromatography columns were purchased from Bio-Rad. All buffers and media 
were prepared using water purified through a Barnstead EasyPure II system (18.2 M?/cm 
resistivity). 
 58 
2.2.2 Cloning: 
All plasmids were prepared using mutagenic coding and non-coding primers (Table 2.1) 
according to the Round-the-Horn procedure [152]. Amplification products were subjected to 
blunt-end ligation with T4 DNA ligase and used to transform E. coli (XL-1 Blue) by heat shock 
according to manufacturer instructions, and transformants were selected on the basis of 
ampicillin resistance. Plasmids from candidate colonies were isolated and evaluated by 
diagnostic restriction digest (Table 2.1) and DNA sequence analysis (Davis Sequencing, Davis, 
CA). Plasmids verified to carry the correct mutations were used to transform E. coli (BL-21-
Gold [DE3] pLysS). 
2.2.3 Expression and purification: 
All expression was carried out in E. coli (BL-21-Gold [DE3] pLysS) cells using liquid 
Luria?Bertani broth supplemented with ampicillin and chloramphenicol with constant agitation. 
Wild-type KatG was over-expressed in a soluble form as previously described [151]. However, 
substituting any one of the four strictly conserved residues resulted in accumulation of target 
protein in inclusion bodies so long as expression was carried out at 37 ?C. Upon adjusting the 
temperature to 18 ?C, a substantial proportion (~ 60%) of expressed protein was produced in a 
soluble state for all KatG
C
 variants. Likewise, R691A, E695A, Y697A, and L690A/R691A 
variants of intact KatG were expressed in a soluble form at appreciable levels (40?60%) at 
18 ?C. Interestingly, the L690A variant of  intact KatG was only observed in inclusion bodies 
even at alternative expression temperatures. In all cases, expression was induced with IPTG once 
cells had reached mid-log phase (OD
600
 = 0.4). At 10 h post-induction, cells were harvested by 
centrifugation, and cell pellets were stored at ? 80 ?C until purification. Expression analysis was  
 59 
 
 
Table 2.1: Primer sets for site-directed mutagenesis of I?-helix. 
Protein: Primer Sequences: 
Restriction 
enzymes 
 
L690A 
 
5?-Phos-CTGCGTGCGAGTACGC-3? 
5?-Phos-TGCGACGAGTAGACAACACAGATCG-3? 
 
 
PmI I 
R691A 5?-Phos-GCGTGCGAGTTACGCAGT-3? 
5?-Phos-GCAGACGAGTAGACAACACAGA-3? 
 
Sac I 
E695A 5?-Phos-GCAGTTACGCAGTAGCGATGCCACG-3? 
5?-Phos-AGCACGCACGCAGACGAGTA-3? 
 
Pst I 
Y697A 5?-Phos-GAGTGCAGCAGTAGCGATGCCACG-3? 
5?-Phos-AGCACGCACGCAGACGAGTA-3? 
 
N/A 
L690A/R691A 5?-Phos-CGCGTGCGAGTTACGCAGT-3? 
5?-Phos-GCGCGACGAGTAGACAACACAG-3? 
 
Sac I 
E695A/Y697A 5?-Phos-GCAGTGCAGCAGTAGCGATGCCACG-3? 
5?-Phos-AGCACGCACGCAGACGAGTA-3? 
 
Pst I 
L690/R691A/ 
E695A/Y697A 
5?-Phos-CGTGCTGCAGTGCGCAGTAG-3? 
5?-Phos-CGCGCGACGCAGTAGACAACAC-3? 
Sac I 
 
 
 
 
 60 
performed using a trichloroacetic acid precipitation technique that we have described elsewhere 
[153]. 
Purification of variants expressed at 18 ?C began by resuspending the cell pellets in 
Bugbuster reagent (Novagen, Madison, WI) in the presence of PMSF (0.1 mM) and benzonase 
nuclease (250 U). The cell lysate was centrifuged, and the supernatant loaded onto a Ni-NTA 
column by recirculating the solution with a flow rate of 1 mL/min through the column overnight. 
Following loading, the column was washed in succession with buffer A (50 mM Tris, pH 8.0), 
buffer B (50 mM phosphate buffer, pH 7.0, 200 mM NaCl), buffer B supplemented with 2 mM 
imidazole, and buffer B supplemented with 20 mM imidazole. The target protein was then eluted 
from the column using buffer B with 50 mM imidazole, and finally, buffer B with 200 mM 
imidazole. Excess imidazole was then removed by gel filtration using a 10DG column from 
BioRad. The eluted fractions were collected and analyzed by SDS-PAGE. Protein-containing 
fractions were combined, concentrated, aliquotted, and stored at ? 80 ?C. Concentrations of 
purified proteins were estimated according to the method of Gill and von Hippel [154]. Even 
though no heme precursors had been added to the expression media, a small portion of the 
purified enzyme already had heme incorporated while inside the cells. 
The variants KatG
N
, E695A/Y697A KatG and L690A/R691A/E695A/Y697A KatG were 
expressed in insoluble inclusion bodies. Purification was carried out as previously described for 
KatG
N
 [142] with the following modifications. After the centrifugation of the cell lysate, the 
insoluble pellet was resuspended using 8 M urea in buffer B. Likewise, the initial washing of the 
loaded Ni-NTA column was carried out using 8 M urea in buffer B supplemented with 10 mM 
imidazole. Elution of the protein from the Ni-NTA column was accomplished with 8 M urea in 
 61 
buffer B supplemented with 400 mM imidazole. The eluent was then pooled and dialyzed against 
buffer B (24?30 h; five changes). With these modifications to the purification procedure, we 
obtained KatG
N
 with a greater capacity for reactivation by KatG
C
 than previously observed [144] 
as determined by visible absorption, MCD, and EPR spectroscopies as well as catalase and 
peroxidase activities. 
2.2.4 Enzyme reconstitution and absorption spectra: 
Due to its poor solubility in neutral and acidic solutions, hemin was dissolved in 0.1 M 
KOH. The concentration of hemin stock solutions was determined by the method of Falk [155]. 
Hemin (0.75 equivalents) was added to wild-type KatG, intact I?-helix variants, and KatG
N
. 
These solutions were allowed to incubate for at least 24 h at 4 ?C. These solutions were then 
centrifuged to remove insoluble unincorporated heme and other debris. The final concentration 
of incorporated hemin was determined by the pyridine hemichrome method of Falk [30]. and 
used to calculate molar absorptivities of the heme absorption spectra of each protein. Spectra for 
the ferrous states of these proteins were obtained by adding a small amount (< 10 mg) solid 
dithionite to the ferric form of each. All spectra were obtained at room temperature on a 
Shimadzu UV-1601 spectrophotometer (Columbia, MD) with a cell path length of 1.0 cm. 
2.2.5 Domain mixing and incubation procedures: 
The concentrations of KatG
N
, KatG
C
, and KatG
C
 variants were estimated based on their 
molar absorptivities as follows: KatG
N
 (?416 = 98 mM
? 1 
cm
? 1
); KatG
C
, L690A KatG
C
, R691A 
KatG
C
, E695A KatG
C
, and L690A/R691A KatG
C
 (?280 = 34 mM
? 1
 cm
? 1
); and Y697A KatG
C
, 
E695A/Y697A KatG
C
, and L690A/R691A/E695A/Y697A KatG
C
 (?280 = 33 mM
? 1 
cm
? 1
). 
 62 
Solutions containing 1:1 ratios of KatG
N
 and KatG
C
 or one of its variants were incubated for 
times ranging from 0 to 96 h in the presence of 50 mM phosphate buffer, pH 7.0, 50 mM NaCl at 
4 ?C. Optimal results were obtained with freshly purified and reconstituted KatG
N 
and freshly 
purified KatG
C
. 
2.2.6 Catalase and peroxidase activity assays: 
Peroxidase activity was evaluated by monitoring the production of the ABTS radical over 
time at 417 nm (34.7 mM
? 1 
cm
? 1
) [156]. All assays were carried out at room temperature in 
50 mM acetate buffer, pH 5.0. Catalase activity was evaluated by monitoring the decrease in 
H
2
O
2
 concentration with time at 240 nm (39.4 M
? 1 
cm
? 1
) [157]. All assays were carried out at 
room temperature in 100 mM phosphate buffer, pH 7.0. Initial velocities (v
o
/[E]
T
) were fit to a 
Michaelis?Menten equation (Eq. (1)) by non-linear regression analysis to determine apparent 
kinetic parameters k
cat
, K
M
, and k
cat
/K
M
. 
!o
[E]T
=
tS
+
                                                                                                           (1) 
Here, [E]
T 
is determined on the basis of the heme content of the protein(s) present. In all 
instances reported here, k
cat 
is the asymptotic maximum catalytic output, the apparent K
M
 simply 
corresponds to the concentration of substrate necessary to produce v
o
/[E]
T
 equal to ? k
cat
, and 
k
cat
/K
M
, the apparent second-order rate constant, represents the efficiency with which the 
substrate is processed and corresponds to the slope of the tangent line to rates obtained at low [S] 
(i.e., <K
M
). When evaluating peroxidase activity as a function of H
2
O
2
 concentration, H
2
O
2
-
dependent inhibition was frequently observed. In order to account for the inhibitory effect and 
make accurate determinations of the apparent parameters k
cat
, K
M
, and k
cat
/K
M
, the data were fit 
 63 
to Eq. (2) derived from a general model for substrate-dependent inhibition[158], where K
I
 is an 
apparent dissociation constant corresponding to inhibition due to the substrate. 
!o
[E]T
=
kcat[S]
KM+
2
I
                                                                                                                   (2) 
 
2.3.7 Circular dichroism spectropolarimetry: 
Far-UV (190?300 nm) circular dichroism (CD) spectra were recorded using a Jasco J-810 
spectropolarimeter (Tokyo, Japan). To minimize buffer interference at wavelengths below 
200 nm, all spectra were recorded using 5 ?M enzyme in the presence of 5 mM phosphate buffer, 
pH 7.0, and Suprasil quartz cells (0.1 mm path length). All spectra were recorded at 23 ?C. 
2.2.8 Magnetic circular dichroism (MCD): 
All spectra of intact wild-type KatG and its variants were obtained using 15 ?M of enzyme. 
Spectra for KatG
N
 alone or incubated with KatG
C
 or one of its variants were obtained using 
10 ?M KatG
N
. All concentrations were determined on the basis of heme content. Spectra (700?
350 nm) were recorded in the presence of 50 mM phosphate buffer, pH 7.0 and 50 mM NaCl at 
23 ?C on a Jasco J-810 spectropolarimeter equipped with a magnetic cell holder (1.4 T) (Tokyo, 
Japan). Baseline subtraction and spectral analysis were carried out using Jasco J-720 software. 
The ferrous states of all heme-containing variants/domain combinations were prepared by adding 
a small amount (< 10 mg) of sodium dithionite to the ferric state. 
 64 
2.2.9 Electron Paramagnetic Resonance (EPR): 
Spectra were recorded using a Bruker EMX instrument equipped with an Oxford ESR 900 
cryostat and ITC temperature controller. Additional sample concentration was performed using 
Amicon Ultra-4 centrifugal devices. The settings for the spectrometer were as follows: 
temperature, 10 K; microwave frequency, 9.38 GHz; microwave power, 0.1 mW; modulation 
amplitude, 10 G; modulation frequency, 100 kHz; time constant, 655.36 ms; conversion time, 
655.36 ms; and receiver gain, 1.0 ? 10
5
. Spin quantification was carried out using the 
Biomolecular EPR Spectroscopy Software package available online [159, 160]. 
2.2.10 GuHCl-mediated protein unfolding: 
The effect of GuHCl on the stability of wtKatG and its I?-helix variants was evaluated by 
two methods. The first method monitored the loss of heme from the active site using a SX.18MV 
(PC upgrade) stopped-flow rapid reaction analyzer (Applied Photophysics, Surrey, UK) 
equipped with diode array detection. One syringe contained 2 ?M wtKatG (or one of its variants) 
and the second syringe contained GuHCl (from 1.5 to 6 M). The kobs values for heme loss were 
determined from the decrease in absorbance at 413.9 nm. All reactions were carried out at 23 ?C 
in 50 mM phosphate pH 7.0. The GuHCl-dependent disruption of the overall secondary 
structural content of KatG and its I? helix variants was monitored by far UV (250?190 nm) CD 
spectroscopy. All reactions were carried out in the presence of 5 mM phosphate buffer, pH 7.0, 
and contained 5 ?M enzyme except for KatG
N
, E695A/Y697A KatG and 
L690A/R691A/E695A/Y697A KatG, which contained 10 ?M protein. Each spectrum was 
corrected using a baseline spectrum obtained using phosphate buffer alone. 
 65 
2.2.11 Temperature-mediated protein unfolding: 
Thermal denaturation of KatG at the active-site level was monitored by visible absorption 
due to the heme prosthetic group. Intact KatG and its I?-helix variants (10 ?M) were incubated 
across a temperature range from 0 to 70 ?C in 50 mM phosphate buffer, pH 7.0. For each 
measurement, enzymes were incubated at the target temperature for 10 min at which point 
spectra were recorded. Due to protein aggregation at high temperature, all spectra were 
background corrected based on their absorbance at 450 nm. 
2.2.12 Calculation of thermodynamic parameters: 
The fraction (?) of unfolded protein at each temperature (T) was calculated from the 
absorbance of the Soret ?max of the native enzyme according to Eq. (3), where A represents the 
absorbance at a given temperature, A
N
 is the absorbance value for the native state, and A
U
 is the 
absorbance of the unfolded state. Similarly from CD data, ? was calculated according to Eq. (4), 
where ?
N
 is the ellipticity at 222 nm for the native folded state, ? is the ellipticity at a given 
GuHCl concentration, and ?
U
 is the ellipticity at 222 nm for the unfolded state. Assuming a two-
state model, where N is the native protein and U is the unfolded protein, the equilibrium constant 
between the unfolded and native state is given by Eq. (5) [161, 162]. This is related to the 
standard free energy of denaturation according to Eq. (6), where R is the universal gas constant 
(8.3145 J mol
? 1 
K
? 1
) and T is the absolute temperature. In unfolding experiments with 
chaotropic reagents performed at 25 ?C (298.13 K), there is generally a linear relationship 
between ?G? and denaturant concentration, allowing for extrapolation [145, 163] to obtain 
?G?
H2O
 (see Eq. (7)). Here ?G?
H2O
 is the value of ?G? in the absence of denaturant and m 
reflects the efficacy of the denaturant for unfolding [145]. 
 66 
? = (A
N
 ? A)/ (A
N
 ? A
U
)                         (3) 
? = (?
N
 ? ?)/(?
N
 ? ?
U
)                              (4) 
K = [U] / [N] = ?/(1 ? ?)                         (5) 
?G? = ? RT lnK                                      (6) 
?G? = ?G?
H2O
 ? m[denaturant]              (7) 
 
2.3 Results: 
2.3.1 Mutagenesis, expression and purification of KatG and KatG
C
 I?-helix variants: 
As confirmed by DNA sequence analysis, we successfully produced constructs for 
expression of seven alanine substitutions involving four intersubunit interface residues: L690, 
R691, E695 and Y697. All I?-helix variants as well as wtKatG and KatG
N
 were successfully 
expressed and isolated. Far-UV CD spectra for I?-helix variants of intact KatG and KatG
C
 were 
indistinguishable from those obtained for wtKatG and unmodified KatG
C
, repectively, indicating 
that substitutions did not disrupt protein secondary structural content of proteins (Fig. 2.1). 
 
2.3.2 Spectroscopic properties of I?-helix KatG variants 
UV-visible absorption spectra (Fig. 2.2) for single substitution variants E695A and Y697A 
KatG were very similar to the wild-type enzyme with Soret band maxima between 406 and 407 
nm and charge transfer transitions around 505 and 640 nm. The spectrum for R691A KatG was 
also very similar except for a more prominent shoulder at 380 nm consistent with a greater 
contribution from pentacoordinate rather than hexacoordinate high-spin heme. By contrast, the 
double (L690A/R691A and E695A/Y697A) and quadruple (L690A/R691A/E695A/Y697A)  
 67 
Figure 2.1A 
190 215 240 265 290
-20000
-10000
0
10000
20000
30000
40000
KatG
C
E695A KatG
C 
R691A KatG
C 
L690A KatG
C 
Y697A KatG
C
L690A/R691A KatG
C
 
E695A/Y697A KatG
C 
wavelength (nm)
[
!
]
 
(
d
e
g
 
c
m
2
 
d
m
o
l
-
1
 
)
 
Figure 2.1B 
Figure 2.1: (A) Far-UV circular dichroism of wtKatG, and I?helix variants. (B) Far-UV 
circular dichroism of C-terminal domain I?helix variants and KatG
C
.  
Spectra were recorded at 23 ?C in 5mM phosphate buffer, pH 7.0 using a 0.5 mm quartz cell. 
 
 
 
190 215 240 265 290
-20000
-10000
0
10000
20000
30000
R691A
L690/R691A
Y697A
E695A
wtKatG
E695A/Y697A
L690A/R691A/
E695A/Y697A
wavelength(nm)
[
!
]
 
(
d
e
g
 
c
m
2
 
d
m
o
l
-
1
)
 68 
 
Figure 2.2: Absorption spectra of ferric wtKatG, its I?-helix substitution variants, and 
KatG
N
.  
Spectra were recorded for wtKatG (a), Y697A (b), E695A (c), R691A (d), L690A/R691A (e), 
E695A/Y697A (f), L690A/R691A/E695A/Y697A (g), and KatG
N
 (h).  
 
 
 
 69 
variants all showed a red shift in the Soret band, increased contributions from the ? and ? bands 
(538 nm and 558 nm respectively) and diminished intensity of the charge transfer transitions. 
The effect was more modest for the L690A/R691A variant but progressively more dramatic for 
E695A/Y697A and the quadruple variant. These results were consistent with a progressively 
greater contribution from hexacoordinate low-spin heme to the spectra.  
 
In order to more effectively differentiate the high-spin and low-spin state contributions of 
these variants, we obtained magnetic circular dichroism (MCD) spectra (Fig. 2.3). Particularly 
with the ferrous oxidation state, the differences between high-spin and low-spin are pronounced 
and easily distinguished. Spectra for the low-spin state shows an intense band with a maximum 
near 550 nm [164]. The dominant feature in the ferrous MCD spectra for all the single 
substitution variants was the B-term features near 440 nm with little if any contribution from the 
A-term feature centered at 557 nm. The A-term became more pronounced and the 440 nm B-
term less pronounced for the double and quadruple variants. These shifts were most subtle for the 
L690A/R691A variant, but spectra for E695A/Y697A KatG and the quadruple variant were 
dominated by features consistent with hexacoordinate low-spin heme. Nevertheless, even with 
the quadruple variant there was a detectable (albeit minor) contribution from high-spin species 
(Fig. 2.3).  
 
 More definitive assessments of the distribution of heme states were obtained from EPR 
spectra (Fig. 2.4). The wild-type and single substitution variants were all dominated by high-spin 
species with little contribution from hexacoordinate low-spin complexes. Further, as is typically 
observed, more than one high-spin complex was detected [142, 165-167]. For L690A/R691A,  
 70 
400 450 500 550 600 650 700
-75
-50
-25
0
25
50
75
100
wavelength (nm)
!
"
/
H
 
(
c
m
-
1
 
M
-
1
 
T
-
1
)
 
Figure 2.3: MCD spectra of ferrous wtKatG, KatG
N 
and I?helix substitution variants.  
Spectra were recorded for 15 ?M wtKatG (bold line), KatG
N
 (solid line), E695A/Y697A (dotted 
line), L690A/R691A/E695A/Y697A (dashed line), Y697A, E695A, R691A, L690A/R691A 
(grey solid lines), in presence of 50 mM phosphate 50 mM NaCl, pH 7.0 using a 5.0 quartz cell 
in a 1.4 T magnetic cell holder. 
 
 
 
 
 
 
 
 71 
 
Figure 2.4: EPR spectra recorded for wtKatG, its I?-helix substitution variants, and KatG
N
.  
Spectra for ferric wtKatG (a), Y697A (b), E695A (c), R691A (d), L690A/R691A (e), 
E695A/Y697A (f), L690A/R691A/E695A/Y697A (g), and KatG
N
 (h) are shown. Spectrometer 
settings were as follows: temperature, 10 K; microwave frequency, 9.38 GHz; microwave power, 
0.1 mW; modulation amplitude, 10 G; modulation frequency, 100 kHz. The g-values 
corresponding to RHS 1 (g
Ax
, g
Ay
, and g
Az
) and RHS 2 (g
Bx
, g
By
, and g
Bz
) are shown as those of 
the hexacoordinate low spin complex (g
x
, g
y
, and g
z
) 
 
 
 72 
E695A/Y697A and the quadruple variant, there was a progressively diminishing contribution 
from the high spin states as well as an increasing contribution from the hexacoordinate low-spin 
complex. To quantify the relative proportions of each species, the spectra were simulated and the 
individual signals were integrated (Table 2.2). The simulations shown used two rhombic high-
spin species (RHS) (RHS 1: g
x
=6.64, g
y
=4.95 and g
z
=1.95; RHS 2: g
x
=5.99, g
y
=5.60 and 
g
z
=1.99) and one low-spin signal (g
x
 =2.92, g
y
 =2.28, g
z
 =1.53). The simulations were also 
carried out using an additional axial high-spin component similar to that described by 
Svistunenko et al. [167], but these did not appreciably change our principal observations. Species 
producing features like RHS 1 have typically been assigned as pentacoordinate high-spin, 
whereas those corresponding to RHS 2 are attributed to a hexacoordinate high-spin state [165-
167]. Wild-type (70%), Y697A (77%) and E695A (56%) were all dominated by RHS 2. The 
R691A variant was dominated by RHS 1 (62%) with a modest contribution from the low-spin 
state (14%). Interestingly, the other variant bearing an R691 substitution L690A/R691A was also 
dominated by RHS 1 (55%), but had a greater contribution from the low-spin state (25%). For 
E695A/Y697A (47%) and the quadruple variant (88%), the most abundant component was the 
hexacoordinate low-spin complex. But for small contribution from RHS 1 (6%) and RHS 2 (6%) 
the latter closely resembled the exclusively hexacoordinate low-spin spectrum of KatG
N
. 
 
2.3.3 Steady-state kinetics of I?-helix KatG variants  
Single substitutions to I?-helix residues caused only modest changes in catalase activity as 
determined by comparing apparent k
cat
 and k
cat
/K
M
 values (Table 2.3). Even the double variant  
 
 
 73 
Table 2.2: Ratios of EPR signals observed for wtKatG, its I?-helix variants, and KatG
N
. 
Protein  % contribution  
 HS 
a
 : LS 
b
 RHS 1 
c
: RHS 2 
d
 RHS 1: RHS 2 : LS 
wtKatG 92 : 8 24 : 76 2 : 70 : 8 
R691A 86 : 14 72 : 28 62 : 24 : 14 
E695A 96 : 4 42 : 58 40 : 56 : 4 
Y697A 93 : 7 18 : 82 16 : 7 : 7 
L690A/R691A 75 : 25 69 : 31 5 : 20 : 25 
E695A/Y697A 53 : 47 57 : 43 30 : 23: 47 
L690A/R691A/E695A/Y697A 12 : 8 53 : 47 6 : 6 : 8 
KatG
N
 0 : 10 0 : 0 0 : 0 : 10 
 
a
 : High Spin; HS = RHS 1+RHS 2 
b
 : Low Spin (g
x
 =2.92 , g
y
 =2.28 , g
z
 =1.53) 
c
 : Rhombic high-spin species 1 (g
x
=6.64, g
y
=4.95 and g
z
=1.95) 
d
 : Rhombic high-spin species 2 (g
x
=5.99, g
y
=5.60 and g
z
=1.99) 
 
 
 
 
 
 
 
 
 74 
Table 2.3: Apparent catalase kinetic parameters for wtKatG and its I?-helix variants 
a 
Protein
  
Parameter
 
 k
cat
 (s
-1
) K
M
 (mM H
2
O
2
) k
cat
/K
M
 (M
-1
 s
-1
) 
wtKatG 11500 ? 200 4.5 ? 0.2 2.5 ? 10
6
 
R691A 6800 ? 300 2.7 ? 0.3 2.5 ? 10
6
 
E695A 8900 ? 500 2.9 ? 0.4 3.1 ? 10
6
 
Y697A 8300 ? 300 1.9 ? 0.2 4.4 ? 10
6
 
L690A/R691A 4800 ? 300 2.7 ? 0.4 1.7 ? 10
6
 
E695A/Y697A 1020 ? 70 1.2 ? 0.3 0.85 ? 10
6
 
L690A/R691/E695A/Y697A 160 ? 30 2.2 ? 1.2 0.07 ? 10
6
 
 
a
 Assay included 20 nM enzyme in 100 mM phosphate buffer, pH 7.0, 23 ?C. 
 
 
 
 
 
 
 
 
 
 
 
 75 
Table 2.4: Apparent peroxidase kinetic parameters for wtKatG and its I?-helix variants
a 
Protein 
Peroxide-dependent peroxidase parameters 
 
k
cat
 (s
-1
) 
K
M 
(mM H
2
O
2
) 
k
cat
 K
M
 (M
-1
 s
-1
) 
K
i
 (mM H
2
O
2
 
) 
wtKatG 
65.0 ? 6.0 
0.180 ? 0.040 
3.60 ? 10
5
 
2.90 ? 1.0 
R691A 105.5 ? 12 0.170 ? 0.040 
6.20 ? 10
5
 
3.35 ? 0.9 
E695A 58.8 ? 5.1 0.13 ? 0.034 
4.41 ? 10
5
 
4.64 ? 1.0 
Y697A 94.7 ? 4.9 0.103 ? 0.019 
9.19 ? 10
5
 
8.51 ? 1.4 
L690A/R691A 6.9 ? 6.9 0.218 ? 0.053 
1.08 ? 10
5
 
5.26 ? 1.3 
E695A/Y697A 7.81 ? 0.7 0.070 ? 0.020 
3.06 ? 10
5
 
7.82 ? 2.1 
L690A/R691A/E695A/Y697A 0.74 ? 0.1 4.70 ? 0.080 
0.02 ? 10
4
 
- 
Protein 
ABTS-dependent peroxidase parameters 
 
k
cat
 (s
-1
) 
K
M
(mM ABTS) 
k
cat
 /K
M
 (M
-1
 s
-1
) 
Ki (mM ABTS ) 
wtKatG 
5.2 ? 1.3 
0.087 ? 0.08 
6.34 ? 10
5
 
- 
R691A 75.5 ? 5.8 0.154 ? 0.025 
4.92 ? 10
5
 
5.96 ? 2.1 
E695A 57.6 ? 5.2 0.078 ? 0.019 
7.38 ? 10
5
 
4.4 ? 1.8 
Y697A 72.0 ? 1.5 0.081 ? 0.07 
8.89 ? 10
5
 
- 
L690A/R691A 46.7 ? 2.6 0.087 ? 0.012 
5.37 ? 10
5
 
8.27 ? 3.2 
E695A/Y697A 6.21 ? 0.2 0.030 ? 0.07 
2.07 ? 10
5
 
- 
L690A/R691A/E695A/Y697A 0.26 ? 0.1 0.07 ? 0.07 
0.37 ? 10
5
 
- 
 
a
 Assay included 20 nM enzyme in 50 mM acetate buffer, pH 5.0 23 ?C. 
 
 
 
 
 76 
L690A/R691A retained about half the catalase activity of the wild-type enzyme. However, the 
apparent k
cat
 values for E695A/Y697A and the quadruple variant were about one and two orders 
of magnitude below the wild-type enzyme, respectively. We observed a similar trend in 
peroxidase activity (Table 2.4). These data corresponded well to the proportion of 
hexacoordinate low-spin heme observed in EPR spectra. 
 
2.3.4 Effect of I?-helix substitution on KatG stability: 
As mentioned in Materials and methods, substituting the strictly conserved residues in the 
I?-helix resulted in expression of protein in insoluble inclusion bodies.  Soluble protein could be 
obtained if expression was carried out at lower temperatures.  This suggested that these residues 
might be central to the proper folding and stability of the enzyme. To investigate this further, we 
evaluated the effect of guanidine hydrochloride (GuHCl) on the secondary structural content of 
KatG and its variants by far-UV circular dichroism. In addition, the sensitivity of the KatG active 
site to disruption by GuHCl or temperature variation was monitored by stopped-flow and UV-
visible absorption, respectively.  All three methods revealed that single substitutions to the I?-
helix had minimal influence on KatG stability.  Disruption of secondary structure by GuHCl was 
indistinguishable for wild-type KatG and the single variants (Fig. 2.5).  Extrapolated free 
energies of unfolding and mid-point unfolding concentrations of GuHCl (C
m
?s) were all similar 
for wild-type KatG, R691A, E695A, Y697A, and even the double variant L690A/R691A (Table 
2.5). At the active site level, these same variants showed only modest increases in the rates of 
GuHCl-mediated heme loss, and thermally induced heme loss was virtually identical for wild-
type KatG, the single variants, and L690A/R691A. Interestingly, in this latter experiment the 
E695A variant did stand out from the group with a T
m
 about 6 ?C lower than the wild-type  
 77 
 
Table 2.5: Parameters GuHCl-mediated unfolding of wtKatG and I?-helix variants. 
Protein  Parameter 
 
?G?
H2O 
(kJ mol 
-1
) 
m
 a
 
(kJ mol
-1
 M
-1
) 
C
m
 b
 
(M) 
k
obs
 
c 
(s
 -1
) 
wt KatG 12.3 ? 0.63 6.86 ? 0.32 1.76 ? 0.03 0.020 ? 0.010 
R691A 1.0 ? 0.6 6.28 ? 0.37 1.70 ? 0.04 0.105 ? 0.04 
E695A 10.2 ? 0.47 5.78 ? 0.28 1.78 ? 0.03 0.134 ? 0.042 
Y697A 11.5 ? 0.58 6.27 ? 0.31 1.72 ? 0.04 0.231 ? 0.010 
L690A/R691A 8.93 ? 0.41 5.52 ? 0.21 1.6 ? 0.51 0.356 ? 0.088 
E695A/Y697A
d f 
12.8 ? 1.73 7.14 ? 0.78 2.3 ? 0.07   8.58 ? 0.036 
E695A/Y697A
e f 
4.9 ? 0.27 4.51 ? 0.30 0.45 ? 0.21   8.58 ? 0.036 
L690A/R691A/E695A/Y697A 
f
 7.47 ? 0.97 25.8 ? 2.8 0.27 ? 0.02   8.47 ? 0.023 
KatG
N 
9.96 ? 1.12
 
51.9 ? 4.8
 
0.19 ? 0.01
 
8.7 ? 0.019
 
 
a 
m, efficacy of GuHCl in unfolding. 
b 
C
m
, GuHCl concentration with K=[U]/[N]=1 
c
 From single exponential fit of plots of absorbance at A
413.9 nm
 in the presence of 2M GuHCl.  
d
 Obtained from the second of two phase of  GuHCl-mediated unfolding.  
e
 Obtained from the first of two phase of  GuHCl-mediated unfolding.  
f
 protein aggregation was observed in presence of <1M GuHCl. 
 
 
 
 
 78 
 
Figure 2.5: GuHCl-mediated unfolding of wtKatG, its I?-helix variants, and KatG
N
 as 
monitored by far-UV CD spectroscopy.  
Relative changes in ellipticity at 222 nm were monitored for wtKatG (?), R691A (?), Y697A 
(?), L690A/R691A (?), E695A (?), E695A/Y697A (?), L690A/R691A/E695A/Y697A (?), and 
KatG
N
 (?). All reactions were carried out using 5 mM phosphate buffer, pH 7.0, at 23 ?C. For 
KatGs indicated by closed symbols, 5 ?M protein was used, and for those indicated by open 
symbols, 10 ?M protein was used. The fraction of the unfolded state, ?, was calculated according 
to Eq. (4). 
 
 
 
 
 79 
 
Figure 2.6: Thermal stability of the heme cavity of wtKatG and its I?-helix substitution 
variants.  
The effect of temperature on heme absorption spectra was evaluated for wtKatG (?), R691A 
(?), Y697A (?), L690A/R691A (?), E695A (?), E695A/Y697A (?), and 
L690A/R691A/E695A/Y697A (?). All temperature profiles were carried out using 100 mM 
phosphate buffer, pH 7.0. Background correction for individual spectra was carried out based on 
absorbance at 450 nm. The fraction of active site-disrupted KatG, ?, was calculated according to 
Eq. (3). 
 
 
 
 
 80 
enzyme (Fig. 2.6).   
 
The E695A/Y697A KatG and quadruple variant were considerably less stable.  The 
secondary structural content of the E695A/Y697A variant showed a two-component response to 
GuHCl. The T
m
 corresponding to the temperature dependent loss of heme for this variant was 
16 ?C lower than the wild-type enzyme, and the rate of heme loss upon reaction with 2M GuHCl 
was over 400 fold faster than wild-type KatG.  The quadruple variants showed an even greater 
loss of stability.  The T
m
 was 24 ?C below wild-type, and the C
m
 with respect to GuHCl 
concentration was over seven fold lower.  The rate constant for GuHCl mediated unfolding was 
also over 400-fold faster. Interestingly, these properties bore a striking resemblance to KatG 
lacking its entire C-terminal domain (i.e., KatG
N
).   
 
2.3.5 Spectroscopic properties of KatG
N
 + KatG
C 
I?-helix substitution variants: 
Expressed as a separate protein, the KatG N-terminal domain (KatG
N
) has neither catalase 
nor peroxidase activity, and the dominant form of heme is a hexacoordinate low-spin complex 
[142]. Introducing the separately expressed and isolated C-terminal domain (KatG
C
) results in 
restructuring of KatG
N
 that is evident from a return of high-spin coordination states typical of 
wild-type enzyme as well as the restoration of catalase and peroxidase activities [144]. 
Specifically, incubation of KatG
N
 with equimolar KatG
C
 at 4 ?C results in a blue shift of the 
Soret band from 415.5 to 410 nm, a decrease in the intensity of the ? and ? bands at 535 and 560 
nm, respectively, and an increase in intensity of charge transfer bands at 505 and 640 nm (Fig.  
2.7).  By EPR, the rhombic signal (g
x
 =2.92, g
y
 =2.28, g
z
 =1.53) corresponding to hexacoordinate 
low-spin heme diminishes in intensity along with a pronounced increase in the two rhombic  
 81 
 
Figure 2.7: UV?visible absorption spectra of KatG
N
 following incubation with KatG
C
 or its 
I?-helix substitution variants.  
Spectra were recorded for the following combinations: KatG
N
 + KatG
C
 (a), KatG
N
 + L690A 
KatG
C
 (b), KatG
N
 + Y697A KatG
C
 (c), KatG
N
 + R691A KatG
C
 (d), KatG
N
 + E695A KatG
C
 (e), 
KatG
N
 + L690A/R691A KatG
C
 (f), KatG
N
 + E695A/Y697A KatG
C
 (g) and KatG
N
 alone (h). All 
reactions contained 10 ?M KatG
N
 and, if present, equimolar KatG
C
 (or variant thereof).  
 
 
 82 
signals mentioned before corresponding to hexacoordinate and pentacoordinate high-spin heme, 
respectively (Fig. 2.8). To varying degrees single substitutions to the I?-helix of KatG
C
 resulted 
in a greater residual proportion of hexacoordinate low-spin states than observed with unmodified 
KatG
C
. The L690A variant most closely resembled unmodified KatG
C
, producing a shift in Soret 
max to 410.5 nm, followed by Y697A (411 nm), R691A (412.5 nm), and E695A (414.5 nm) 
with the latter showing little ability to alter the coordination state of the heme in the KatG
N
 active 
site (Fig. 2.7).  A nearly identical trend was observed by EPR spectroscopy (Fig. 2.8). Individual 
EPR signals were integrated and spectra simulated to quantify the relative proportions of high-
spin and low-spin species as described previously (Table 2.6). Of the single variants, E695A 
KatG
C
 had the most diminished capacity to induce a return of high spin heme species in KatG
N
, 
with over 50% of the KatG
N
 heme remaining in its low-spin form. R691A and Y697 KatG
C
 left 
35 ? 40% KatG
N
 heme in its low-spin state.  Meanwhile, L690A KatG
C
 showed a capacity only 
slightly below unmodified KatG
C
 to induce a return of high spin heme species in KatG
N
.  As 
evaluated by either UV- vis or EPR, KatG
C
 double variants L690A/R691A and especially 
E695A/Y697A lost almost all ability to alter the KatG
N 
heme coordination state.  
 
2.3.6 Kinetic study of KatG
N
 + KatG
C 
I?-helix substitution variants: 
It has been previously shown that incubation of KatG
N
 with KatG
C
 not only produces spectral 
shifts due to the change in heme environment, but also the return of both catalase and peroxidase 
activities [144]. Steady-state kinetic parameters for recovered catalase (Table 2.7) and 
peroxidase (Table 2.8) activities follow very closely the trends observed in spectroscopic 
experiments.  Any substitution to the I?-helix diminished the ability of KatG
C
 to recover catalase 
and peroxidase activity.  The effect was most pronounced with the E695A substitution and  
 83 
 
 
Figure 2.8: EPR spectra of KatG
N
 following incubation with KatG
C
 or its I?-helix 
substitution variants.  
Spectra were recorded for the following combinations: KatG
N
 + KatG
C
 (a), KatG
N
 + L690A 
KatG
C
 (b), KatG
N
 + Y697A KatG
C
 (c), KatG
N
 + R691A KatG
C
 (d), KatG
N
 + E695AKatG
C
 (e), 
KatG
N
 + L690A/R691AKatG
C
 (f), KatG
N
 + E695A/Y697A KatG
C
 (g) and KatG
N
 alone (h). The 
g-values corresponding to RHS 1 (g
Ax
, g
Ay
, and g
Az
) and RHS 2 (g
Bx
, g
By
, and g
Bz
) are shown as 
those of the hexacoordinate low spin complex (g
x
, g
y
, and g
z
). 
 
 84 
Table 2.6: Ratios of EPR signals observed following KatG
N
 incubation with KatG
C
 or its I?helix 
substitution variants 
Protein  % contribution  
 HS
a
 : LS
b
 RHS 1
c
: RHS 2
d
 RHS 1: RHS 2 : LS 
KatG
N
+KatG
C
 94 : 6 63 : 37 59 : 35 : 6 
KatG
N
+L690A KatG
C
 8 : 12 60 : 40 53 : 35 : 12 
KatG
N
+R691A KatG
C
 62 : 38 82 : 18 51 : 1 : 38 
KatG
N
+E695A KatG
C
 45 : 5 60 : 40 27 : 18 : 5 
KatG
N
+Y697A KatG
C
 63 : 37 71 : 29 4 : 19 : 37 
KatG
N
+L690A/R691AKatG
C
 19 : 81 75 : 25 14 : 5 : 81 
KatG
N
+E695A/Y697AKatG
C
 3 : 97 4 : 56 1 : 2 : 97 
KatG
N
 0 : 10 0 : 0 0 : 0 : 10 
 
a
 : High Spin; HS = RHS 1+RHS 2 
b
 : Low Spin (g
x
 =2.92 , g
y
 =2.28 , g
z
 =1.53) 
c
 : Rhombic high-spin species 1 (g
x
=6.64, g
y
=4.95 and g
z
=1.95) 
d
 : Rhombic high-spin species 2 (g
x
=5.99, g
y
=5.60 and g
z
=1.99) 
 
 
 
 
 
 
 
 
 
 85 
progressively less for R690A, Y697A, and least for L690A.  Likewise, multiple substitutions to 
I?-helix residues completely abrogated the ability of KatG
C
 to restore catalytic function to the N-
terminal domain.    
 
I?-helix substitutions also diminished the rate of KatG
N
 active site restructuring by KatG
C
 
(Fig. 2.9, Table 2.7). Here the same trend was observed as in the spectroscopic and steady-state 
activity experiments.  The most rapid rate of KatG
N
 activity recovery, of course, was facilitated 
by the unmodified KatG
C
.  Any substitution to the I?-helix diminished the rate with the E695A 
producing the largest effect followed by R691A, Y697A, and finally L690A.  As described 
above, two or more substitutions to the I?-helix produced KatG
C
 with no capacity to reactivate 
KatG
N
.  
 
2.4 Discussion: 
All catalase-peroxidases have a two-domain structure which appears to have arisen from 
the gene duplication and fusion of a predecessor peroxidase [118]. Since that event, the N- 
terminal domain has either gained or retained the capacity for heme-dependent reduction and 
oxidation of H
2
O
2
 as well as a classical, albeit limited, peroxidatic activity. In contrast, the C-
terminal domain has lost the ability to bind heme or catalyze any reaction. Nevertheless, the C-
terminal domain is an integral component of catalase-peroxidase structure and catalysis [142, 
144, 145], but its precise roles in KatG function are still unknown.  
 
One hypothesis has been that the C-terminal domain serves as a platform to support the 
folding of the N-terminal domain [143] that in the absence of the C-terminal domain, the N- 
 86 
 
 
 
Figure 2.9: Time course for recovery of KatG
N
 catalase activity upon addition of equimolar 
KatG
c
 or one of its I?helix substitution variants.  
Activity measurements were made for the following combinations: KatG
N
 + KatG
C
 (?), 
KatG
N
 + L690A KatG
C
 (?), KatG
N
 + Y697A KatG
C
 (?), KatG
N
 + R691A KatG
C
 (?), and 
KatG
N
 + E695A KatG
C
 (?). Incubations were carried out at 4 ?C in 100 mM phosphate buffer, 
pH 7.0. Assays of the activity contained 50 nM KatG
N
 and KatG
C
 (or variant thereof), 10 mM 
H
2
O
2
, and 100 mM phosphate buffer, pH 7.0. Assays were all performed at 23 ?C. 
 
 
 
 
 
 87 
 
 
Table 2.7: KatG
N
 reactivation k
obs
 and apparent catalase parameters for reactivated enzyme. 
Protein Catalase cycle parameter
a
 Reactivation
b
 
 
k
cat 
(s
-1
) 
K
M 
(mM) 
k
cat
/K
M 
(10
6
 M
-1
 s
-1
) 
k
obs
 
(10
-5
 s
-1
) 
KatG
N
+KatG
C
 400 ? 350 2.4 ? 0.6 1.7 3.83 ? 0.56 
KatG
N
+L690A KatG
C
 350 ? 340 4.0 ? 0.9 0.9 2.25 ? 0.18 
KatG
N
+R691A KatG
C
 20 ? 20 3.6 ? 0.3 0.6 1.36 ? 0.08 
KatG
N
+E695A KatG
C
 1050 ? 170 6.0 ? 1.9 0.2 0.81 ? 0.1 
KatG
N
+Y697A KatG
C
 270 ? 180 3.3 ? 0.5 0.8 1.53 ? 0.26 
KatG
N
+L690A/R691A KatG
C
 ND
 c
 ND ND ND 
KatG
N
+E695A/Y697A KatG
C
 ND ND ND ND 
 
a
 All assays included 50 nM enzyme in 100 mM phosphate buffer, pH 7.0, at 23 ?C. 
b
 Incubation conditions for reactivation: 100 mM phosphate buffer, pH 7.0, at 4 ?C. 
c 
Not
 
determined. 
 
 
 
 
 
 
 
 
 88 
 
Table 2.8: Apparent Peroxidase parameters for reactivated KatG
N
 by KatG
C
 or its I? helix 
variants.
a 
Protein Peroxide-dependent parameters 
  
 
k
cat
 (s
-1
) 
K
M 
(mM H
2
O
2
) 
k
cat
/K
M
 (M
-1
 s
-1
) 
Ki (mM H
2
O
2 
) 
KatG
N
+KatG
C
 48.7 ? 1.0 0.068 ? 0.07 7.20 ? 10
5
 - 
KatG
N
+ L690A KatG
C
 16.0 ? 0.4 0.075 ? 0.08 2.13 ? 10
5
 - 
KatG
N
+ R690A KatG
C
 2.8 ? 0.5 0.07 ? 0.07 2.98 ? 10
5
 - 
KatG
N
+ E695A KatG
C
 6.10 ? 0.1 0.039 ? 0.03 1.54 ? 10
5
 - 
KatG
N
+ Y697A KatG
C
 10.8 ? 1.4 0.027 ? 0.020 4.07 ? 10
5
 - 
KatG
N
+L690A/R691AKatG
C
 ND ND ND - 
KatG
N
+E695A/Y697AKatG
C
 ND ND ND - 
Protein ABTS-dependent parameters 
  
 
k
cat
 (s
-1
) K
M
(mM ABTS) k
cat
/K
M
 (M
-1
 s
-1
) Ki (mM ABTS ) 
KatG
N
+KatG
C
 56.2 ? 2.9 0.423 ? 0.061 1.32 ? 10
5
 9.81 ? 1.4 
KatG
N
+ L690A KatG
C
 14.0 ? 0.8 0.16 ? 0.024 1.20 ? 10
5
 13.5 ? 3.2 
KatG
N
+ R690A KatG
C
 2.0 ? 1.7 0.086 ? 0.021 2.56 ? 10
5
 7.67 ? 1.9 
KatG
N
+ E695A KatG
C
 5.60 ? 0.4 0.042 ? 0.010 1.3 ? 10
5
 7.30 ? 1.5 
KatG
N
+ Y697A KatG
C
 14.0 ? 0.6 0.095 ? 0.014 1.47 ? 10
5
 9.72 ? 1.5 
KatG
N
+L690A/R691AKatG
C
 ND ND ND - 
KatG
N
+E695A/Y697AKatG
C
 ND ND ND - 
 
a
 Assay included 50 nM enzyme in 50 mM acetate buffer, pH 5.0 at 23 ?C. 
 
 89 
terminal domain alone has no apparent activity, and its heme is in a hexacoordinate low-spin 
ferric state [142]. This is remedied by the reintroduction of the separately expressed and isolated 
C-terminal domain (KatG
C
) which restores a correct coordination state as well as both catalase 
and peroxidase activities [144]. 
 
Visual inspection of known KatG structures reveals substantial intrasubunit as well as 
intersubunit interactions between N- and C-terminal domains. One prominent example is the I?-
helix which appears at the interface between the two domains and contacts several elements of 
the N-terminal domain. As such, the I?-helix may serve as a central component of a hypothetical 
folding platform provided by the C-terminal domain [119, 143].  
 
One aspect of our approach to test this hypothesis was to evaluate the effect of substitution 
to the I?-helix on the ability of KatG
C
 to reactivate KatG
N
. Indeed, our data indicate that even 
single substitutions to the I?-helix substantially limit the ability of KatG
C
 in this respect.  All of 
the single variants showed a decrease in the rate and extent of KatG
N
 reactivation. With two 
substitutions to the I?-helix, KatG
C
 become completely incapable of directing the restructuring of 
KatG
N
 reactivation.  Clearly, the I?-helix is a critical component in this process. 
 
Because preparation of KatG
N
 involves denatured purification, refolding, and 
reconstitution, it is reasonable to argue that the reactivation of KatG
N
 by KatG
C
 may not 
adequately reveal the extent to which the C-terminal domain directs folding events and structural 
adjustments to the N-terminal domain during expression in vivo. In light of this, it is compelling 
that we observed similar trends and phenomena for I?-helix substitutions to intact KatG. All of 
 90 
the intact KatG variants showed a greater tendency toward expression in inclusion bodies. In 
addition, the double and quadruple variants showed a substantially increased propensity to form 
the hexacoordinate low-spin state as well as clear compromises to active site and overall protein 
structural stability. Indeed, the quadruple variant showed properties nearly identical to KatG
N
 in 
all respects even though it retained what appeared to be a correctly folded C-terminal domain. 
We hasten to add that the effects of I?-helix substitutions were more subtle for the intact protein 
than KatG
C
. Where most single substitutions produced notable defects in the restructuring 
activity of KatG
C
, those same substitutions to intact KatG showed only modest influences on 
stability, coordination state, and catalytic function. 
 
There are extensive hydrogen-bonded networks between the I?-helix of the C-terminal 
domain and its intersubunit N-terminal domain partner [119-122] (Fig. 2.10A). In particular, 
E695 (Fig. 2.10B) forms an H-bond with conserved W159 at the N-terminal end of the d-helix. 
Further, in conjunction with the R126 side chain, E695 typically coordinates two water 
molecules. Interestingly, this same R126 guanidium group, a strictly conserved residue of the BC 
interhelical connecting loop, makes several additional polar contacts including the carbonyl 
oxygen of residue 188, and through a well-ordered water molecule, the invariant D187 
carboxylate, and the carbonyl oxygens of residues 122 and 123. The former pair is located on the 
DE helical connecting loop, and the latter pair is found on the BC loop. Interestingly, in Class II 
peroxidases (e.g. manganese peroxidase), which depend on Ca
2 +
 to maintain the heme 
coordination environment that supports activity, the relevant calcium-binding site is established 
 91 
in part by this same portion of the BC loop [168, 169]. More broadly, this network connects the 
I?-helix with several core components of the peroxidase fold of the N-terminal domain. 
Interestingly, E695 also appears to participate in another network that also includes R691 
and R144 (Fig. 2.10B). Though less well defined, these strictly conserved arginines form 
additional hydrogen bonds with the backbone carbonyl of Q293 (Met in M. tuberculosis KatG), a 
residue from large loop 2 (LL2). In addition, the side chain of amino acid 293 forms tight 
interactions with hydrophobic residues from the I?-helix. Notably, our EPR spectra of R691A 
showed a greater contribution from the RHS 1 component, the pentacoordinate high-spin species, 
than either wild-type or the other single variants. In an interesting corroboration of those results, 
we have observed that KatG lacking LL2 (or Q293-containing portions thereof) shows nearly 
identical spectroscopic signatures and catalytic properties [130]. 
Unlike E695 and R691, L690 and Y697 side chains are located such that they make contact 
with both inter- and intrasubunit N-terminal domain partners (Fig. 2.10 A and C). The lesser 
impact of L690 and Y697 may owe to the fact that both residues are in contact with the N-
terminal end of the N-terminal domain rather than its core structures like the BC loop, C helix or 
LL2, all of which appear to be crucial for protein active site folding and catalytic ability. 
In summary, at first glance the KatG C-terminal domain has all the appearances of a 
vestigial structure. It bears the general structure of a peroxidase and even the correct helical 
framework to support an active site, but this domain shows no ability to bind heme or catalyze 
any reaction. The closest relatives of KatG, cytochrome c peroxidase and ascorbate peroxidase, 
have no C-terminal domain. Indeed, it is proposed that the single domain peroxidases were 
 92 
 
 
Figure 2.10: Interfaces between the KatG N-terminal domain and I?-helix from the C-
terminal domain.  
The secondary structural context of the I?-helix from each KatG subunit is shown in panel A. A 
magnified view of the interactions of E695 and R691 is shown in panel B, and that 
corresponding to L690 and Y697 is shown in panel C. Coordinates are from the structure of 
Burkholderia pseudomallei KatG (PDB:1MWV). Numbering reflects Escherichia coli KatG 
sequence. Secondary structures in green are from subunit A and those in blue are from subunit B. 
Dotted lines indicate inferred polar contacts. Red spheres represent water molecules.  
 
 93 
obtained from the deletion of one domain from a two-domain predecessor [92]. Nevertheless, the 
C-terminal domain is a universal component of KatGs, and without it, the enzyme does not 
function. Such a large structure could conceivably have several roles in KatG function, but one 
proposal has been that it serves as a folding platform for the active site-bearing N-terminal 
domain [143]. We have capitalized on the unique ability of KatG
C
 to restructure and reactivate 
an otherwise inactive KatG
N
 to show that the C-terminal domain's I?-helix may be a central 
component of a platform to facilitate folding events and/or conformational adjustments in the N-
terminal domain. Though less dramatic, our results with intact KatG corroborate those obtained 
with the separately prepared domains. In particular, substituting four strongly conserved residues 
in I?-helix resulted in a protein whose properties were nearly identical to KatG
N
 even though it 
retained what appeared to be a correctly folded C-terminal domain. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 94 
 
 
 
 
 
CHAPTER THREE 
 
 
FACILITATING KATG INTERDOMAIN COMMUNICATION: CONNECTING THE 
REMOTE I?-HELIX TO ACTIVE SITE CONFORMATION, STABILITY, AND 
CATALYTIC FUNCTION 
 
3.1 Introduction: 
Two of the most common enzymes in nature for disposing of H
2
O
2
 are heme-dependent 
peroxidases and catalases. A unique family of enzymes that show both catalase and peroxidase 
activities (KatGs) are found among archaea, eubacteria, fungi and protists [93, 170, 171]. The 
typical catalase and peroxidase cycles both start with the heterolytic reduction of H
2
O
2
 along 
with the oxidation of the ferric heme to a ferryl porphyrin/protein radical intermediate 
(compound I). In the typical catalase cycle, compound I is then reduced back to resting enzyme 
by a second equivalent of H
2
O
2
, generating O
2
 and H
2
O. Conversely, the peroxidase returns to 
the ferric state by two sequential one-electron steps using an exogenous reducing substrate. The 
first reduction produces one equivalent of substrate radical and either a ferryl or a ferric-
hydroxy/protein radical intermediate (compound II), and the second reductive step produces the 
resting enzyme and a second equivalent of substrate radical. 
 
Catalase-peroxidases belong to class I of a superfamily formerly designated as the fungal, 
plant and bacterial peroxidases [118]. Driven in part by the dramatic increase in sequence data 
for all types of peroxidases, the nomenclature has recently been adjusted to designate this 
 95 
superfamily as the peroxidase-catalases [21]. KatGs are observed either as homodimers or 
homotetramers, where each subunit is composed of two homologous domains. Neither domain 
bears any resemblance to monofunctional catalases, but both have the topological structure of 
monofunctional peroxidases. Only the N-terminal domain retains an active site, and it is 
superimposable with those of typical peroxidases [119, 120, 122, 146]. The C-terminal domain is 
? 30 ? from the active site. It is conserved across all catalase-peroxidases, and its topological 
similarity to peroxidases suggests that it originated from a gene-duplication and fusion event 
[118, 128, 141]. However, the roles of this second, apparently ?inactive?, domain in catalysis are 
not immediately obvious. It bears a striking resemblance to monofunctional peroxidases, but it 
neither binds heme nor catalyzes any discernable reaction [119, 120, 122, 143, 146]. It has been 
suggested that the C-terminal domain might serve as a platform to direct proper folding of the N-
terminal domain and facilitate dimerization [143, 144]. Our own results support this assertion 
and further suggest that the I?-helix is a necessary structure for directing structural adjustments in 
the adjacent N-terminal domain active site (See Chapter 2) [172]. We noted that E695 was 
particularly important as substitution of this strictly conserved residue was destructive to KatG 
active site integrity and profoundly disrupted the ability of KatG
C
 to induce structural adjustment 
in its KatG
N
 partner. 
 
Visual inspection of the crystal structures of KatG indicates that E695 forms an H-bond 
with conserved W159 found at the N-terminal end of the D-helix. It also, through two well-
ordered water molecules, forms substantial contacts with R126, a residue found on the BC 
helical connecting loop [119, 120, 122, 146]. These are core structures of the peroxidase fold, 
and may account for the substantial effect of E695 substitution on KatG active site, conformation, 
 96 
stability and function. Therefore, E695 located 25 ? from the active site may through these 
interactions modulate the N-terminal domain active site conformation and functionality. 
 
        To address these interactions and their impacts on the KatG active site. We produced single 
and double alanine substitution variants of intact KatG, KatG
N
 and KatG
C
. These were 
expressed, isolated and characterized. Here, we report the effect of these substitutions not only 
on active site structure, stability and bifunctionality in intact KatG, but also on KatG
N
 
reactivation by KatG
C
. The spectroscopic and kinetic results demonstrate that strictly conserved 
R126 and W159 are critical for KatG active site structure, stability and catalytic function. 
Through its interaction with these residues, E695 of the I?-helix modulates the active site 
conformation and function. 
 
3.2 Materials and Methods: 
3.2.1 Materials: 
Hydrogen peroxide (30%), imidazole, hemin, ampicillin, chloramphenicol, sodium 
dithionite, phenylmethylsulfonyl fluoride (PMSF), 2,2?-azino-bis (3-ethylbenzthiazoline-6-
sulfonate) (ABTS), and guanidine hydrochloride (GuHCl) were purchased from Sigma (St. 
Louis, MO). Isopropyl-?-d-thiogalactopyranoside (IPTG), urea, mono- and dibasic sodium 
phosphate, acetic acid, and sodium acetate were obtained from Fisher (Pittsburgh, PA). 
Bugbuster and benzonase were purchased from Novagen (Madison, WI). All restriction enzymes 
were purchased from New England Biolabs (Beverly, MA). All oligonucleotide primers were 
purchased from Invitrogen (Carlsbad, CA). All E. coli strains (BL-21-Gold [DE3] pLysS and 
XL-1 Blue), Pfu polymerase, and T4 DNA ligase were obtained from Agilent (La Jolla, CA). 
 97 
Nickel-nitrilotriacetic acid (Ni-NTA) resin was purchased from Qiagen (Valencia, CA). 
Desalting 10DG chromatography columns were purchased from Bio-Rad. All buffers and media 
were prepared using water purified through a Barnstead EasyPure II system (18.2 M?/cm 
resistivity). 
3.2.2 Cloning:  
All plasmids were prepared using mutagenic coding and non-coding primers (Table 3.1) 
according to the Round-the-Horn procedure [152]. Amplification products were subjected to 
blunt-end ligation with T4 DNA ligase and used to transform E. coli (XL-1 Blue) by heat shock 
according to manufacturer instructions, and transformants were selected on the basis of 
ampicillin resistance. Plasmids from candidate colonies were isolated and evaluated by 
diagnostic restriction digest (Table 3.1) and DNA sequence analysis (Davis Sequencing, Davis, 
CA). Plasmids verified to carry the correct mutations were used to transform E. coli (BL-21-
Gold [DE3] pLysS). 
3.2.3 Expression and Puri?cation:  
All expression was carried out in E. coli (BL-21-Gold [DE3] pLysS) using liquid Luria?
Bertani broth supplemented with ampicillin and chloramphenicol with constant agitation. Wild-
type KatG was over-expressed in a soluble form as previously described [151]. However, 
variants with substitutions to E695 and W159 were expressed in inclusion bodies at 37 ?C. In 
order to express soluble forms, we reduced the temperature for expression to 18 ?C. The KatG
C
 
was over-expressed in a soluble form with a six His-tag at 37 ?C as previously described [144]. 
In all other respects, expression and purification was carried out as previous described (Chapter 
2). Concentrations of purified enzyme were estimated according to the method of Gill and von  
 
 98 
 
Table 3.1: Primer sets for site-directed mutagenesis. 
Protein: Primer Sequences: 
Restriction 
enzymes 
E695A 
 
5?-Phos-GCAGTTACGCAGTAGCGATGCCACG-3? 
5?-Phos-AGCACGCACGCAGACGAGTA-3? 
 
Pst I 
E695Q 
5?-Phos-GCAGTATACGCAGTAGCGATGCACG-3? 
5?-Phos-GCACGCACGCAGACGAGTAG-3? 
 
Ac I 
R126A 
5?-Phos-CTTGCACGCTGACTCTGCG-3? 
5?-Phos-CTGCTGACACGACCGCGC-3? 
 
Hind II 
W159A 
5?-Phos-GTCAGAATCTCGCGCGACTGT-3? 
5?-Phos-CATATTCTGTTGATGCACACAGCGACGC-3? 
Eag I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 99 
Hippel [154]. KatG
N
, R126AKatG
N
 and W159AKatG
N
 were expressed in inclusion bodies and 
were purified under denaturing conditions as described in Chapter 2. 
 
3.2.4 Enzyme reconstitution:  
Enzyme reconstitution followed the procedures described in Chapter 2. In order to 
minimize effects of free or adventitiously bound hemin to spectral or kinetic measurement after 
domain incubation, reconstitution of KatG
N 
was carried out by adding 0.75 equivalents of hemin 
to enzyme. These solutions were allowed to incubate for at least 24 h at 4 ?C. These solutions 
were then centrifuged to remove insoluble unincorporated heme and other debris. Molar 
absorptivities were as follows: W159A KatG
N
 and KatG
N 
(?
416
 = 98 mM
? 1
 cm
? 1
); R126A KatG
N 
(?
416
 = 113 mM
? 1
 cm
? 1
) KatG
C
, E695Q KatG
C
, and E695A KatG
C
 (?
280
 = 34 mM
? 1
 cm
? 1
); 
Solutions containing 1:1 ratios of KatG
N
 and KatG
C
 or one of its variants were incubated for 
times ranging from 0 to 96 h in the presence of 50 mM phosphate buffer, pH 7.0, 50 mM NaCl at 
4 ?C. Optimal results were obtained with freshly purified and reconstituted KatG
N
 and freshly 
purified KatG
C
. 
 
3.2.5 Absorption spectra:   
The final concentration of incorporated hemin was determined by the pyridine hemichrome 
method of Falk [155] and used to calculate molar absorptivities of the heme absorption spectra of 
each protein. Spectra for the ferrous states of these proteins were obtained by adding a small 
amount (< 10 mg) solid dithionite to the ferric form of each. All spectra were obtained at room 
temperature on a Shimadzu UV-1601 spectrophotometer (Columbia, MD) with a cell path length 
of 1.0 cm. 
 100 
3.2.6 Catalase and peroxidase activity assays:  
Peroxidase activity was evaluated by monitoring the production of the ABTS radical over 
time at 417 nm (34.7 mM
? 1 
cm
? 1
) [156]. All assays were carried out at room temperature in 
50 mM acetate buffer, pH 5.0. Catalase activity was evaluated by monitoring the decrease in 
H
2
O
2
 concentration with time at 240 nm (39.4 M
? 1 
cm
? 1
) [157]. All assays were carried out at 
room temperature in 100 mM phosphate buffer, pH 7.0. Initial velocities (v
o
/[E]
T
) were fit to a 
Michaelis?Menten equation (Eq. (1)) by non-linear regression analysis to determine apparent 
kinetic parameters k
cat
, K
M
, and k
cat
/K
M
. 
!o
[E]T
=
tS
+
 
 
                                                                                  (1) 
Here, [E]
T
 is determined on the basis of the heme content of the protein(s) present. In all 
instances reported here, k
cat
 is the asymptotic maximum catalytic output, the apparent K
M
 simply 
corresponds to the concentration of substrate necessary to produce v
o
/[E]
T
 equal to ? k
cat
, and 
k
cat
/K
M
, the apparent second-order rate constant, represents the efficiency with which the 
substrate is processed and corresponds to the slope of the tangent line to rates obtained at low [S] 
(i.e., <K
M
). When evaluating peroxidase activity as a function of H
2
O
2
 concentration, H
2
O
2
-
dependent inhibition was frequently observed. In order to account for the inhibitory effect and 
make accurate determinations of the apparent parameters k
cat
, K
M
, and k
cat
/K
M
, the data were fit 
to Eq. (2) derived from a general model for substrate-dependent inhibition [173], where K
I
 is an 
apparent dissociation constant corresponding to inhibition due to the substrate.
 
!o
[E]T
=
kcat[S]
KM+
2
I
                                                                                              (2) 
 
 101 
3.2.7 Circular Dichroism Spectropolarimetry:  
Far-UV (190?300 nm) circular dichroism (CD) spectra were recorded using a Jasco J-810 
spectropolarimeter (Tokyo, Japan). To minimize buffer interference at wavelengths below 
200 nm, all spectra were recorded using 5 ?M enzyme in the presence of 5 mM phosphate buffer, 
pH 7.0, and Suprasil quartz cells (0.1 mm path length). All spectra were recorded at 23 ?C. 
 
3.2.8 Magnetic Circular Dichroism:  
All spectra of intact wild-type KatG and its variants were obtained using 15 ?M enzyme. 
Spectra for KatG
N
 alone or incubated with KatG
C
 or one of its variants were obtained using 
10 ?M KatG
N
. All concentrations were determined on the basis of heme content (see above). 
Spectra (700?350 nm) were recorded in presence of 50 mM phosphate buffer, pH 7.0 and 50 mM 
NaCl at 23 ?C on a Jasco J-810 spectropolarimeter equipped with a magnetic cell holder (1.4 T) 
(Tokyo, Japan). Baseline subtraction and spectral analysis were carried out using Jasco J-720 
software. The ferrous states of all heme-containing variants/domain combinations were prepared 
by adding a small amount (< 10 mg) of sodium dithionite to the ferric state. 
 
3.2.9 Electron Paramagnetic Resonance:  
Spectra were recorded using a Bruker EMX instrument equipped with an Oxford ESR 900 
cryostat and ITC temperature controller. Additional sample concentration was performed using 
Amicon Ultra-4 centrifugal devices. The settings for the spectrometer were as follows: 
temperature, 10 K; microwave frequency, 9.38 GHz; microwave power, 0.1 mW; modulation 
amplitude, 10 G; modulation frequency, 100 kHz; time constant, 655.36 ms; conversion time, 
655.36 ms; and receiver gain, 1.0 ? 10
5
. Spin quantification was carried out using the 
 102 
Biomolecular EPR Spectroscopy Software package available online [159, 160]. 
 
3.2.10 Evaluation of protein stability:  
The effect of GuHCl on the stability of wtKatG and its intersubunit interface variants was 
evaluated by two methods similar as previous described in Chapter 2 except unfolding study by 
CD spectroscopy was carried out using 5 ?M KatG (or variants). Thermal denaturation of KatG 
at the active site level was monitored by visible absorption due to the heme prosthetic group as 
described in Chapter 2. Due to protein aggregation at high temperature, all spectra were 
background corrected based on their absorbance at 450 nm. Calculation of thermodynamic 
parameters was carried out according to equations 3-7 as described in Chapter 2. 
 
3.3 Results: 
3.3.1 Mutagenesis, expression and purification of KatG and its intersubunit variants:  
As confirmed by DNA sequence analysis, we successfully produced constructs for 
expression of alanine substitutions involving three strictly conserved residues along the 
intersubunit interface: R126, W159 and E695. As described in Materials and methods, all the 
single substitution variants, as well as the double variants R126A/E695A and W159A/E695A, 
were successfully expressed and isolated. There was no observable change to secondary 
structural content as determined on the basis of the circular dichroism spectra of the variants 
when compared to the wild-type enzyme (Fig. 3.1). 
 
 
 103 
 
190 200 210 220 230 240 250 260
-15000
-7500
0
7500
15000
22500
E695Q
R126A
W159A
W159A/E695A
R126A/E695A
wtKatG 
E695A
Wavelength (nm)
[
!
]
 
(
d
e
g
 
c
m
2
 
d
m
o
l
-
1
)
 
Figure 3.1: Far-UV circular dichroism of wtKatG, and its E695, R126 and W159 
substitution variants. 
 
 
 
 
 
 
 
 
 
 104 
360 400 440 525 600 675
407
506
642
a
b
c
d
e
g
f
Wavelength (nm)
330360390420450480
0
20
40
60
80
100
1 week
48h
407nm 409nm
Wavelength (nm)
!
 
(
m
M
-
1
 
c
m
-
1
)
 
Figure 3.2: UV-visible absorption spectra of ferric wtKatG and its E695, R126 and W159 
substitution variants.  
All spectra were recorded for the proteins wtKatG (a), E695Q (b), E695A (c), R126A (d), 
R126A/E695A (e), W159A (f), W159A/E695A (g) in their native ferric state in 100 mM 
phosphate buffer, pH 7.0, in a 1.0 cm quartz cell at 23 ?C. The inset shows a comparison of 
spectra of R126A/E695A taken after 48 h (solid line) and taken after 1 week (dashed line) 
reconstituted by hemin. 
 
 
 105 
3.3.2 Spectroscopic properties of R126A, W159A and E695Q: 
UV-visible absorption spectra for all the single and double substitution variants were 
nearly identical to the wild-type enzyme with Soret band maxima between 407 and 407.5 nm and 
charge transfer transitions around 506 and 640 nm, indicating the active site was dominated by 
high-spin ferric heme species (Fig. 3.2). With the exception of the R126A/E695A variants, all 
spectra recorded following storage for one week at 4 ?C were unchanged. For R126A/E695A 
there was a red shift in the Soret band to 409 nm along with increased contributions from the ? 
and ? bands at 538 nm and 558 nm, respectively, and diminished intensity of the charge transfer 
transitions. These features are consistent with increased contribution from hexacoordinate low-
spin heme (Fig. 3.2-inset).  
 
In order to more effectively differentiate the high- and low-spin contributions of these 
variants, magnetic circular dichroism (MCD) spectra were recorded and analyzed. The 
distinction between the high- and low-spin states is particularly stark in the ferrous rather than 
ferric oxidation state (as described in Chapter 2). The dominant feature in the ferrous MCD 
spectra for all the substitution variants was the B-term near 440 nm, similar to the wild-type 
enzyme. Only the R126A/E695A double variant showed a more intense A-term centered near 
560 nm along with a diminished B-term at 440 nm. These data are consistent with a greater 
contribution from hexacoordinate low-spin heme as observed in UV-vis spectra (Fig. 3.3). 
 
Electron paramagnetic resonance (EPR) spectroscopy was used to further evaluate the 
influence of substitutions on heme iron environment in the native ferric state (Fig. 3.4). All the 
substitution variants showed EPR spectra similar to the wild-type, especially in that high-spin  
 106 
400 500 600 700
-25
0
25
50
75
Wavelength (nm)
!
"
/
H
 
(
c
m
-
1
 
M
-
1
 
T
-
1
)
 
Figure 3.3: MCD spectra of ferrous wtKatG and its E695, R126 and W159 substitution 
variants.  
Spectra were recorded for 15 ?M wtKatG (bold line black), W159A (solid line black), E695A 
(bold line grey), E695Q (solid line grey), W159A/E695A (dotted line), R126A/E695A (dashed 
line) in presence of 50 mM phosphate 50 mM NaCl pH 7.0 using a 5.0 quartz cell in a 1.4 T 
magnetic cell holder. 
 
 
 
 
 
 
 
 107 
1000 2000 3000 4000 5000
g
BX
=6.64
g
Ax
=5.99
g
BY
=4.95 g
Az
=1.99
g
Bz
=1.95
g
Bx
=5.60
g=3.55 g=2.00
a
b
c
d
e
g
f
Field [G]
 
Figure 3.4: EPR spectra of wtKatG
 
and its E695, R126 and W159 substitution variants.  
Spectra for ferric wtKatG (a), E695Q (b), E695A (c), R126A (d), R126A/E695A (e), W159A (f), 
W159A/E695A (g) are shown. Spectrometer settings were as follows: temperature, 10 K; 
microwave frequency, 9.38 GHz; microwave power, 0.1 mW; modulation amplitude, 10 G; 
modulation frequency, 100 kHz. The g-values corresponding to RHS 1 (g
Ax
, g
Ay
, and g
Az
) and 
RHS 2 (g
Bx
, g
By
, and g
Bz
) are shown as those of the HALS-type low spin complex (g=3.55, 
g=2.0). 
 
 108 
species dominated over low-spin states. However, the spectra showed greater detail in the 
distribution of the various high- and low-spin states present.  Similar to the wild-type enzyme, 
both E695A and E695Q KatG showed a prominent contribution from the rhombic high-spin 
signal we designate as RHS 2 (g
x
=5.99 g
y
=5.60 g
y
=1.99) with relatively small contributions from 
the second rhombic signal, RHS 1 (g
x
=6.64, g
y
=5.60 and g
z
=1.95). The same features were 
observed with the R126A variant, but the distribution among the high spin states was distinct 
with a greater contribution from the RHS 1 component, and less from the RHS 2 signal. The 
E695A-bearing double variants (R126A/E695A and W159A/E695A) showed a similar 
preference for the RHS 1 species, but unexpectedly, a highly anisotropic low spin (HALS)-type 
EPR signal, characterized by features at g = 3.55, g = 2.00 was also detectable with these 
variants. This latter spectral feature is proposed to arise due to strain on the heme iron, at the near 
perpendicular orientation of the imidazole side chains of two histidine ligands [174-176]. To 
quantify the relative proportions of each species, the spectra were simulated and the individual 
signals were integrated (Table 3.2).  
 
3.3.3 Steady-state kinetics of R126A, W159A and E595Q:  
Single substitutions to intersubunit interface residues resulted in minimal changes in 
catalase activity as determined by comparing apparent k
cat
 and k
cat
/K
M
 values. Even with the 
double variants, only moderate effects were observed with respect to k
cat
, and the apparent 
second-order rate constants were unchanged compared to wtKatG (Table 3.3). Likewise, any 
substitution to E695, R126 or W159 resulted in only small decreases in peroxidase k
cat
 values 
with respect to both H
2
O
2
 and the electron donor substrate (ABTS in these studies). The most 
detrimental substitution with respect to peroxidase activity was R126A, which showed an order  
 109 
 
Table 3.2: Ratios of EPR signals observed in wtKatG and its E695, R126 and W159 substitution 
variants. 
Protein   % contribution  
 HS
a
 : LS
b
 RHS 1
c
: RHS 2
d
 RHS 1: RHS 2 : LS RHS 1: RHS 2 : LS (H) 
e
 
wtKatG 92 : 8 24 : 76 2 : 70 : 8 ND 
f
 
E695Q 97 : 3 60 : 40 58 : 39 : 3 ND 
E695A 96 : 4 42 : 58 40 : 56 : 4 ND 
R126A 96 : 4 78 : 2 75 : 21 : 4 ND 
W159A 95 : 5 56 : 4 53 : 42 : 5 ND 
R126A/E695A ND 69 : 31 ND 34 : 15 : 51 
W159A/E695A ND 80 : 20 ND 40 : 10 : 50 
 
a
 High spin; HS = RHS 1 + RHS 2. 
b
 Low spin (g
x
 = 2.92, g
y
 = 2.28, g
z
 = 1.53). 
c
 Rhombic high-spin species1 (g
x
=6.64, g
y
=4.95 and g
z
=1.95). 
d
 Rhombic high-spin species 2 (g
x
=5.99, g
y
=5.60 and g
z
=1.99). 
e
 HALS-type low-spin species (g=3.55, g=2.00). 
f
 Not determined. 
 
 
 
 
 
 
 
 110 
 
Table 3.3: Apparent catalase kinetic parameters for wtKatG and its E695, R126 and W159 
substitution variants
 a 
Protein
  
Parameter
 
 
 k
cat
 (s
-1
) K
M
 (mM H
2
O
2
) k
cat
/K
M
 (M
-1
 s
-1
) 
wtKatG     11500 ? 200 4.5 ? 0.2 2.5 ? 10
6
 
E695Q 15300 ? 1000 5.3 ? 0.7 2.9 ? 10
6
 
E695A 9000 ? 500 2.9 ? 0.4 3.1 ? 10
6
 
R126A 14000 ? 1200 8.0 ? 1.3 1.7 ? 10
6
 
W159A 13500 ? 1200 2.6 ? 0.6 5.2 ? 10
6
 
R126A/E695A 7400 ? 600 1.9 ? 0.4 4.0 ? 10
6
 
     W159A/E695A 5900 ? 300 1.7 ? 0.3 3.5 ? 10
6
 
 
a
 Assay included 20 nM enzyme in 100 mM phosphate buffer, pH 7.0, 23 ?C. 
 
 
 
 
 
 
 
 
 
 
 
 
 111 
Table 3.4: Apparent peroxidase kinetic parameters for wtKatG and its E695, R126 and W159 
substitution variants
 a 
Protein 
Peroxide-dependent peroxidase parameters 
 
k
cat
 (s
-1
) 
K
M 
(mM H
2
O
2
) 
k
cat
/K
M
 (M
-1
 s
-1
) 
Ki (mM H
2
O
2
 
) 
wtKatG 
72.0 ? 5.1 
0.27 ? 0.06 
3.60 ?10
5
 
27.3 ? 10.4 
E695Q 67.9 ? 2.7 0.39 ? 0.04 
1.74 ?10
5
 
10.1 ? 1.0 
E695A 58.8 ? 5.1 0.13 ? 0.03 
4.41 ?10
5
 
4.6 ? 1.0 
R126A 19.0 ? 1.5 0.34 ? 0.07 
0.5 ?10
5
 
7.4 ? 1.6 
W159A 56.3 ? 4.5 0.43 ? 0.08 
1.08 ?10
5
 
10.1 ? 2.0 
R126A/E695A 3.4 ? 2.4 0.17 ? 0.04 
1.96 ?10
5
 
7.7 ? 1.6 
W159A/E695A 40.0 ? 2.0 0.16 ? 0.02 
2.50 ?10
5
 
6.7 ? 0.9 
Protein 
ABTS-dependent peroxidase parameters 
 
k
cat
 (s
-1
) 
K
M
(mM ABTS) 
k
cat
/K
M
 (M
-1
 s
-1
) 
Ki (mM ABTS ) 
wtKatG 
76.0 ? 4.0 
0.134 ? 0.023 
5.6 ?10
5
 
- 
E695Q 40.0 ? 1.0 0.076 ? 0.07 
5.26 ?10
5
 
- 
E695A 57.6 ? 5.2 0.078 ? 0.019 
7.38 ?10
5
 
- 
R126A 39.6 ? 8.9 0.090 ? 0.041 
4.40 ?10
5
 
0.51 ? 0.2 
W159A 35.8 ? 1.2 0.067 ? 0.09 
5.34 ?10
5
 
- 
R126A/E695A 26.9 ? 0.9 0.035 ? 0.06 
7.68 ?10
5
 
- 
W159A/E695A 34.4 ? 0.7 0.065 ? 0.07 
5.29 ?10
5
 
- 
 
a
 Assay included 20 nM enzyme in in 50 mM acetate buffer, pH 5.0 at 23 ?C 
 
 
 
 112 
of magnitude decrease in k
cat
/K
M
 with respect to H
2
O
2
. Interestingly, R126A was also the only 
variant to show ABTS- dependent inhibition (Table 3.4). However, the peroxidase activity was 
not diminished to the same degree for the double substitution variant R126A/E695A, where the 
apparent k
cat
/K
M 
was nearly 50% of that of wtKatG. 
 
3.3.4 The influence of substitution of R126, W159 and E695 on KatG stability:  
As mentioned in Material and methods, substitution of strictly conserved residues E695 
and W159 resulted in the expression of protein in insoluble inclusion bodies using standard 
expression conditions (i.e., 37 ?C). However, soluble protein was obtained if expression was 
carried out at lower temperatures (i.e., 18 ?C). This suggested that these residues might be central 
to the proper folding and stability of the enzyme. To investigate this further, we evaluated the 
effect of guanidine hydrochloride (GuHCl) on the secondary structural content of KatG and its 
variants by far-UV CD. In addition, the sensitivity of the KatG active site to disruption with 
temperature variation or GuHCl concentration were monitored by UV?visible absorption and 
stopped-flow spectrophotometry, respectively. 
 
 
From far-UV CD data, free energies of unfolding were extrapolated and mid-point 
unfolding concentration (C
m
) of GuHCl were determined (Table 3.5). They were all similar for 
wild-type KatG, E695A, and E695Q.  Even the W159 substitution variants showed only modest 
decreases in the C
m
 for GuHCl-mediated unfolding in comparison to the wild-type enzyme (Fig. 
3.5). In contrast, far greater sensitivity of the active site to GuHCl-mediated disruption was 
observed, with about an order magnitude faster rate constants for heme loss in the presence of 2 
M GuHCl for these variants as compared to those measured for the wild-type enzyme. Likewise,  
 113 
 
0 1 2 3
0.0
0.2
0.4
0.6
0.8
1.0
WtKatG
E695Q
E695A 
[GuHCl] (M)
!
0 1 2 3
0.0
0.2
0.4
0.6
0.8
1.0
Wt KatG
W159A/
E695A
W159A
[GuHCl] (M)
!
0 1 2 3
0.0
0.2
0.4
0.6
0.8
1.0
WtKatG
R126A/
E695A
R126A 
[GuHCl] (M)
!
 
Figure 3.5: GuHCl-mediated unfolding of wtKatG and E695, R126 and W159 substitution 
variants as monitored by far-UV circular dichroism (CD) spectroscopy.  
 
 
 
 
 
 
 114 
Table 3.5: Parameters of GuHCl-mediated unfolding wtKatG and E695, R126 and W159 
substitution variants. 
Protein 
Parameter 
?G?
H2O 
(kJ mol 
-1
) 
m
 a
 
(kJ mol
-1
 M
-1
) 
C
m
 b
 
(M) 
k
obs
 
c 
(s
 -1
) 
wtKatG 12.3 ? 0.63 6.86 ? 0.32 1.76 ? 0.03 0.020 ? 0.01 
E695A 10.9 ? 0.37 7.0 ? 0.23 1.78 ? 0.03 0.130 ? 0.01 
E695Q 12.5 ? 0.4 6.16 ? 0.2 1.71 ? 0.02 0.139 ? 0.030 
R126A 10.9 ? 0.59 6.25 ? 0.32 1.74 ? 0.04 0.312 ? 0.024 
W159A 10.7 ? 0.45 6.53 ? 0.24 1.56 ? 0.03 0.10 ? 0.01 
R126A/E695A
 
12.0 ? 0.43 7.02 ? 0.24 1.70 ? 0.02 0.604 ? 0.150 
W159A/E695A
 
1.0 ? 0.50 7.10 ? 0.29 1.48 ? 0.02 0.16 ? 0.010 
KatG
N
 9.63 ? 1.27 51.9 ? 4.8 0.19 ? 0.01 8.7 ? 0.019 
 
a 
m, efficacy of GuHCl in unfolding. 
b 
C
m
, GuHCl concentration with K=[U]/[N]=1 
c
 From single exponential fit of A
413.9nm
 decrease in the presence of 2 M GuHCl 
 
 
 
 
 
 
 115 
40 50 60
0.00
0.25
0.50
0.75
1.00
Temperature (?C)
!
 
Figure 3.6: Thermal stability of the heme cavity of wtKatG and E695, R126 and W159 
substitution variants as monitored by UV-visible absorption due to the heme retention.  
Enzymes wtKatG (?), E695A (?), W159A (?), R126A (?), E695Q (?), W159A/E695A (?), 
R126A/E695A (?) were in 100 mM phosphate buffer, pH 7.0 in a 1.0 cm quartz cell. All spectra 
backgrounds were calibrated from 350 nm to 450 nm. ? was calculated based on derived 
equation (1), ? = (A
N soret
 /A
380nm
? A 
soret 
/A
380nm
)/ (A
N soret
 /A
380nm
? A
U soret 
/A
380nm
).  
 
 
 
 
 
 
 116 
evaluation of thermally induced heme loss showed that any substitution to E695, R126 or W159 
diminished active site stability compared to wtKatG. The midpoint temperatures of transition 
(T
m
) were 6 - 8 ?C lower for the single substitution variants and 9 - 10 ?C lower for the double 
substitution variants (Fig. 3.6). 
 
3.3.5 Effect of E695, W159 and E695 substitution on the spectroscopic properties of KatG
C
 
reactivated KatG
N
: 
The stand-alone N-terminal domain of KatG (KatG
N
) has neither catalase nor peroxidase 
activity and the dominant form of heme is a hexacoordinate low-spin complex [142]. Introducing 
the separately expressed and isolated C-terminal domain (KatG
C
) results in restructuring of 
KatG
N
 as evident from a return of high-spin coordination states typical of wild-type enzyme as 
well as the restoration of catalase and peroxidase activities [144].  Specifically, incubation of 
KatG
N
 with equimolar KatG
C
 at 4 ?C results in a blue shift of the Soret band from 415.5 to 410.5 
nm and a decrease in the intensity of the ? and ? bands at 535 and 560 nm, respectively. Along 
with this there is an increase in intensity of charge transfer bands at 505 and 640 nm.  By EPR, 
the typical rhombic signal of the hexacoordinate low-spin state (g
x
 =2.92, g
y
 =2.28, g
z
 =1.53) 
diminishes in intensity along with a pronounced increase in the RHS 2 and RHS 1 signals 
corresponding to hexacoordinate and pentacoordinate high-spin heme states, respectively [144].  
Substitution of E695 in KatG
C
 or R126 or W159 in KatG
N
 resulted in a greater residual 
proportion of hexacoordinate low-spin states after domain incubation than observed with the 
recombined unmodified domains. The R126A variant most resembled unmodified KatG
N
, 
producing a shift in ?
max
 for the Soret band to 411.5 nm; however, W159A (413.5 nm) showed 
little ability to alter the coordination state of the heme. Interestingly, any substitution of the E695 
 117 
side chain in KatG
C
 whether the more dramatic E695A or the more conservative E695Q 
substitution limited its ability to return of high-spin heme states to the KatG
N
 active site, 
producing the maxima of 414.5 nm and 414 nm, respectively. With two substitutions, the ability 
to change the active site heme coordination state was almost completely abrogated (Fig. 3.7). 
 
A nearly identical trend was observed by EPR spectroscopy (Fig. 3.8). As before the 
relative proportions of high-spin and low-spin species were estimated by integration and 
simulation of the individual EPR spectra (Table 3.6). Any single substitution to R126 and E695 
limited the return of KatG
N
 to active high-spin heme states. The most detrimental substitution 
was E695A, which showed the greatest proportion of low-spin species followed by E695Q and 
R126A. We also noticed that substitution of W159 resulted in a protein with poor heme binding 
property as shown from EPR spectra. Its activation by KatG
C
 or its variants showed similar EPR 
features distinct from that of KatG
N
. By UV-vis and EPR spectroscopy, the double variants 
R126A/E695A and W159A/E695A showed little ability to alter the N-terminal domain active 
site
 
heme coordination state. 
 
3.3.6 Effect of E695, W159 and E695 substitution on the catalytic function of KatG
C
 
reactivated KatG
N
: 
As we have observed previously, both catalase and peroxidase activities reemerge 
concomitant with the spectral changes associated with incubation of KatG
N
 with KatG
C 
[144]
.
 
The steady-state kinetic parameters for catalase and peroxidase activities following KatG
N
 
reactivation are shown in Table 3.7 and Table 3.8, respectively. Any substitution to E695, R126 
or W159 diminished recovery of catalase and peroxidase activities.  The effect was most  
 118 
360 400 440 525 600 675
5 !
410.5
526
638.5
414.5
535
414
532.5
411.5
527
415
415.5
538
535.5
558
558
414
414
413.5
535
535
533.5
a
b
c
d
e
f
g
h
i
Wavelength (nm)
 
Figure 3.7: UV-visible absorption spectra of KatG
N 
(or its R126A, W159A substitution 
variants) following incubation with KatG
C
 (or its E695A, E695Q substitution variants).  
Spectra were recorded for the following combinations: KatG
N
+KatG
C
 (a), KatG
N
+E695AKatG
C
 
(b), KatG
N
+E695QKatG
C
 (c), R126AKatG
N
+KatG
C
 (d), R126AKatG
N
+E695AKatG
C
 (e), KatG
N
 
(f), W159AKatG
N 
(g), W159AKatG
N
+KatG
C
 (h) and W159AKatG
N
+E695A KatG
C
 (i). All 
reactions contained 10 ?M KatG
N
 (or variant thereof) and, if present, equimolar KatG
C
 (or 
variant thereof). All incubations were carried out for 96 h at 4 ?C in the presence of 100 mM 
phosphate buffer, pH 7.0. Spectra were recorded using the same buffer at 23 ?C 
 119 
1000 2000 3000 4000 5000
g
AX
= 5.99
g
BX
=6.64
g
BY
=4.95
g=2.92 g=2.28
g
Az
=1.99
g
Bz
=1.95
g=1.53
a
b
c
d
e
f
g
AY
=5.60
g
Field [G]
i
h
 
Figure 3.8: EPR spectra of KatG
N 
(or its R126A, W159A substitution variants) following 
incubation with KatG
C
 (or its E695A, E695Q substitution variants).  
Spectra were recorded for the following combinations: KatG
N
+KatG
C
 (a), KatG
N
+E695A KatG
C
 
(b), KatG
N
+E695QKatG
C
 (c), R126AKatG
N
+KatG
C
 (d), R126AKatG
N
+E695AKatG
C
 (e), KatG
N 
(f), W159AKatG
N 
(g), W159AKatG
N 
+KatG
C
 (h), W159AKatG
N
+E695AKatG
C 
(i). The g-values 
corresponding to RHS 1 (g
Ax
, g
Ay
, and g
Az
) and RHS 2 (g
Bx
, g
By
, and g
Bz
) are shown as those of 
the hexacoordinate low spin complex (g
x
, g
y
, and g
z
). All incubations were carried out as 
described for Fig. 3.7, and spectrometer settings were as described for Fig. 3.4.  
 120 
Table 3.6: Ratios of EPR signals observed following KatG
N
 (or its R126A, W159A substitution 
variants) following incubation with KatG
C
 (or its E695A, E695Q substitution variants) 
Protein  % Contribution  
 HS
a
 : LS
b
 RHS 1
c
: RHS 2
d
 RHS 1: RHS 2 : LS 
KatG
N
+KatG
C
 94 : 6 63 : 37 59 : 35 : 6 
KatG
N
+E695A KatG
C
 45 : 5 60 : 40 27 : 18 : 5 
KatG
N
+E695QKatG
C
 64 : 36 71 : 29 45 : 18 : 36 
R126AKatG
N
+KatG
C
 67 : 3 75 : 25 50 : 17 : 33 
R126AKatG
N
+E695AKatG
C
 15 : 85 50 : 50 7 : 7 : 85 
KatG
N
 0 : 10 0 : 0 0 : 0 : 10 
 
a
 High Spin; HS = RHS 1+RHS 2 
b
 Low Spin (g
x
 =2.92 , g
y
 =2.28 , g
z
 =1.53) 
c
 Rhombic high-spin species 1 (g
x
=6.64, g
y
=4.95 and g
z
=1.95) 
d
 Rhombic high-spin species 2 (g
x
=5.99, g
y
=5.60 and g
z
=1.99) 
 
 
 
 
 
 
 
 
 
 
 
 121 
pronounced with W159A substitution and progressively less for E695A, R126A, and least for 
E695Q. The R126A and E695A substitution appeared to be additive as the R126A KatG
N
 + 
E695A KatG
C
 combination, recovered only half the catalase activity of KatG
N 
+ E695AKatG
C
; 
and W159A KatG
N
 + E695A KatG
C
 combination completely lost the ability to restore catalytic 
function to the N-terminal domain. Notably, the trend of reactivation of peroxidase activity was 
not the same as that of catalase activity. Any substitution to these residues limited recovery of 
peroxidase activity. This effect was more dramatic upon substituting the residues (R126 and 
W159) in the stand-alone N-terminal domain (Table 3.8). 
 
Any substitution to E695, R126 and W159 diminished the first order rate constant 
associated with KatG
N
 functional recovery. The largest effect was observed with the W159A 
variant followed by E695Q, E695A, and finally R126A (Fig. 3.9, Table 3.7). Interestingly, the 
E695Q substitution had the smallest effect on the amplitude of catalase activity recovery, but the 
rate constant was one order magnitude below that observed for the unmodified KatG
C
. With two 
substitutions (i.e, R126AKatG
N 
+ E695AKatG
C
), the rate constant for activity recovery was 
decreased by an order of magnitude. As observed above, W159A KatG
N
 + E695A KatG
C
 
showed little capacity to restructure the N-terminal domain active site. Thus, it was not 
surprising to find little, if any, ability to recover either catalase or peroxidase activity. 
 
3.5 Discussion: 
3.5.1: Intrasubunit interactions vs. intersubunit interactions: 
The C-terminal domain of KatG is a structure found in all catalase-peroxidases. It appears 
to have arisen from gene duplication and fusion of a predecessor peroxidase [118]. The  
 122 
Table 3.7: Apparent catalase kinetic parameters for KatG
N
 (or its R126A, W159A substitution 
variants) following incubation with KatG
C
 (or its E695A, E695Q substitution variants)
 a
 
Protein Catalase cycle parameter
a
 Reactivation
a,b
 
 
k
cat 
(s
-1
) 
K
M 
(mM H
2
O
2
) 
k
cat
/K
M
 
(M
-1
 s
-1
) 
k
obs
 
(10
-5
 s
-1
) 
KatG
N
+KatG
C
 400 ? 350 2.4 ? 0.6 1.7 ? 10
6
 3.83 ? 0.56 
KatG
N
+E695A KatG
C
 100 ? 20 6.0 ? 1.9 0.2 ? 10
6
 0.81 ? 0.1 
KatG
N
+E695Q KatG
C
 260 ? 90 2.2 ? 0.2 1.8 ? 10
6
 0.5 ? 0.07 
R126AKatG
N
+KatG
C
 170 ? 90 1.7 ? 0.3 1.0 ? 10
6
 2.35 ? 0.30 
R126AKatG
N
+E695AKatG
C
 50 ? 40 1.3 ? 0.3 0.4 ? 10
6
 0.54 ? 0.12 
W159AKatG
N
+KatG
C
 230 ? 10 2.5 ? 0.4 0.1 ? 10
6
 0.36 ? 0.14 
W159AKatG
N
+E695AKatG
C
 ND
 c
 ND ND ND 
 
a
 All assays included 50 nM enzyme in 100 mM phosphate buffer, pH 7.0, at 23 ?C. 
b
 Incubation conditions for reactivation: 100 mM phosphate buffer, pH 7.0, at 4 ?C. 
c 
Not
 
determined. 
 
 
 
 
 
 
 
 
 123 
Table 3.8: Apparent peroxidase kinetic parameters for KatG
N
 (or its R126A, W159A 
substitution variants) following incubation with KatG
C
 (or its E695A, E695Q substitution 
variants)
 a 
Protein          Peroxide-dependent peroxidase parameters 
 
k
cat 
(s
-1
) K
M 
(mM H
2
O
2
) k
cat
/K
M
 (M
-1
 s
-1
) K
I 
(mM H
2
O
2
 
) 
KatG
N
+KatG
C
 56.3 ? 2.9 0.42 ? 0.06 1.34 ? 10
5
 9.80 ? 1.4 
KatG
N
+E695AKatG
C
 5.60 ? 0.4 0.04 ? 0.01 1.40 ? 10
5
 7.30 ? 1.5 
KatG
N
+E695QKatG
C
 16.3 ? 2.1 0.10 ? 0.03 4.41 ? 10
5
 1.0 ? 4.7 
R126AKatG
N
+KatG
C
 3.60 ? 0.2 0.09 ? 0.02 0.04 ? 10
5
 10.9 ? 2.3 
W159AKatG
N
+KatG
C
 0.50 ? 0.1 0.03 ? 0.01 0.02 ? 10
5
 27.2 ? 7.2 
R126AKatG
N
+E695AKatG
C
 2.30 ? 0.1 0.08 ? 0.01 0.03 ? 10
5
 9.30 ? 1.4 
W159AKatG
N
+E695AKatG
C
 ND ND ND - 
Protein ABTS-dependent peroxidase parameters 
 
k
cat
 (s
-1
) K
M
(mM ABTS) k
cat
/K
M
 (M
-1
 S
-1
) K
I
 (mM ABTS ) 
KatG
N
+KatG
C
 48.7 ? 1.1 0.068 ? 0.07 6.96 ? 10
5
 - 
KatG
N
+E695AKatG
C
 6.10 ? 0.1 0.039 ? 0.03 1.56 ? 10
5
 - 
KatG
N
+E695QKatG
C
 17.7 ? 0.4 0.043 ? 0.04 4.1 ? 10
5
 - 
R126AKatG
N
+KatG
C
 6.40 ? 0.4 0.014 ? 0.02 4.57 ? 10
5
 0.231 ? 0.04 
W159AKatG
N
+KatG
C
 0.50 ? 0.1 0.032 ? 0.07 0.15 ? 10
5
 - 
R126AKatG
N
+E695AKatG
C
 5.0 ? 1.9 0.089 ? 0.053 0.56 ? 10
5
 0.14 ? 0.07 
W159AKatG
N
+E695AKatG
C
 ND ND ND - 
 
a
 Assay included 20 nM enzyme in in 50 mM acetate buffer, pH 5.0 at 23 ?C 
 
 
 
 124 
0.0 20.0 40.0 60.0
0
1000
2000
3000
4000
time (h)
v
o
/
[
E
]
T
 
(
s
-
1
)
 
Figure 3.9: Time course for recovery of KatG
N
 (or its R126A, W159A substitution variants)
 
catalase activity upon addition of equimolar KatG
C
 (or its E695A, E695Q substitution 
variants).  
Activity measurements were made for the following combinations: KatG
N
+KatG
C
 (?), 
KatG
N
+E695QKatG
C
 (?), KatG
N
+E695AKatG
C
 (?), R126AKatG
N
+KatG
C
 (?), 
W159AKatG
N
+KatG
C
 (?), R126AKatG
N
+E695AKatG
C
 (?), W159AKatG
N
+E695AKatG
C
 (?). 
Incubations were carried out at 4 ?C in 100 mM phosphate buffer, pH 7.0. Assays of the activity 
contained 50 nM KatG
N 
(or variant thereof) and KatG
C
 (or variant thereof), 10 mM H
2
O
2
, and 
100 mM phosphate buffer, pH 7.0. Assays were all performed at 23 ?C. 
 
 
 
 
 
 125 
functional adaption to this event appears to have resulted in an N-terminal domain with the 
ability to oxidize and reduce hydrogen peroxide and remain peroxidatically active. In contrast, 
the C- terminal domain has lost the ability to even bind heme, let alone catalyze any reaction 
with it. Because it is clear that the C-terminal domain does not bring an activity of its own, it is 
reasonable to suppose that it may address the needs of the N-terminal domain through several 
points of contact by which it may direct conformational adjustments, maintain the active site 
coordination environment and fine tune for bifunctional activity [144]. This is verified by visual 
inspection of the crystal structures for KatG enzyme that show substantial interactions between 
the two domains [119, 120, 122, 146], and supported further by the effect not only of deletion of 
the C-terminal domain, but also its reintroduction as a separately expressed and isolated protein 
[142, 144].  
 
Interestingly, the effect of removing the C-terminal domain from KatG produces very 
similar results as removal of calcium from fungal secretory peroxidases (i.e., class II enzymes 
like lignin peroxidase).  Specifically, there is a shift in heme coordination from pentacoordinate 
high-spin to hexacoordinate low-spin. In the class II enzymes, calcium is coordinated above the 
B-helix, and its loss due to temperature increase or disruption of its coordination environment by 
mutagenesis allows the distal histidine (a residue found on the B-helix) to coordinate the heme 
iron (Fig. 3.12) [81, 169, 177-179].  It is striking that the class III peroxidases (e.g., horseradish 
peroxidase) contain a calcium ion in essentially the same position.  However, in these enzymes 
the metal iron coordination site is further stabilized by a disulfide bond where the two cysteine 
residues straddle aspartate residues that serve to coordinate the calcium [65, 180]. None of the 
class I enzymes, including KatG contain Ca
2+
 nor is there a disulfide bond at the analogous 
 126 
position. However, in the case of KatG there are extensive intrasubunit contacts between the BC 
loop of the N-terminal domain and the B?C? loop from the C-terminal domain. Along the two-
domain interface, residues contribute to two prominent hydrogen-bonded networks (Tyr 111-Asp 
482-Arg 479 and Arg 117-Asp 597) both of which are highly conserved. These interactions 
would appear to have the same capacity to stabilize the B-helix and prevent the distal histidine 
from entering the coordination sphere of the heme iron (Fig. 3.10). However, previous studies 
from the Goodwin laboratory have shown that disruption of these hydrogen-bonded networks is 
sufficient neither to replicate the effects observed in KatG upon removal of its C-terminal 
domain nor those in class II peroxidase upon removal calcium [181]. These data tend to suggest 
that the C-terminal domain may direct the active site structural adjustment through some other 
points of contact with N-terminal domain. 
 
As revealed in Chapter 2, the I?-helix which appears at the intersubunit interface plays an 
essential role in stabilizing the N-terminal domain in its active conformation, as multiple 
substitutions to this structure produced substantial increases in the population of hexacoordinate 
low-spin heme and diminished enzyme activity, similar to that observed with N-terminal domain 
on its own (KatG
N
). Indeed, multiple substitutions to the I?-helix produced a KatG with 
spectroscopic and kinetic properties, as well as thermodynamic stability, virtually 
indistinguishable from KatG
N
. We noted that substitution of E695 was especially destructive to 
KatG active site integrity; the stability at the active site level was substantially compromised as 
observed by the thermally-induced loss of active site heme.  
 
3.5.2: Connecting the I?-helix to the active site: 
 127 
 
Figure 3.10: Interactions between the N-terminal domain BC loop and the B?C? loop from 
the intrasubunit C-terminal domain. 
Coordinates from the structure of Mycobacterium tuberculosis KatG were used (PDB: 2CCA). 
Numbering reflects E. coli KatG sequence. Dashes indicate the inferred polar contacts. 
 
 
 
 
 
 
 
 128 
Although the I?-helix clearly has an essential role in maintaining the conformation and 
integrity of active site, the mechanisms by which this structure connects to the active site and 
controls its conformation are not clear. Visual inspection of crystal structures shows that E695 
makes contacts with several core structures of N-terminal domain with connections to the active 
site. One of these is an apparent H-bond with strictly conserved W159 where the indole nitrogen 
appears to be a H-bond donor to the E695 carboxylate. Further, E695 coordinates one or two 
water molecules in conjunction with the R126 side chain between two domains [119, 120, 122, 
146]. 
 
At first glance, replacement of W159 with alanine did not appear to produce any notable 
defect in KatG structure or function as evaluated by its nearly identical spectral and kinetic 
properties to wtKatG. Even in terms of global protein structural stability, this substitution only 
resulted in modest losses as evaluated by effects of denaturing agents on overall secondary 
structural content.  In contrast, the protein stability at the active site level was substantially 
compromised as observed by thermally induced loss of active site heme similar to what we 
observed in E695A. This was further corroborated by the analogous variant in KatG
N
, which 
retained little if any ability to respond to active site restructuring by KatG
C
. Interestingly, 
substitution of R126 with alanine produced results highly similar to the W159A variant where 
the most obvious detriment to the function of the enzyme was the stability of its active site to 
temperature and denaturants. Taken together, the impacts of single substitutions on KatG were 
most obvious in the stability of the active site, indicating these residues are the key connections 
through which the remote I?-helix controls the active site integrity and conformation. 
 
 129 
Visual inspection of crystal structures of KatGs shows that the R126 guanidinium group 
anchors another hydrogen-bonded network through another well-ordered water molecule and 
strictly conserved D187 (Fig. 3.11). This network includes the backbone carbonyl oxygen of Arg 
122 (Gly in M.tuberculosis KatG), the backbone of which further associates with another water 
molecule that appears to stabilize the BC loop located above the B-helix bearing the distal His 
(Fig. 3.11). These well-ordered water molecules which are observed in all structures of KatG 
enzymes provide a potential connection from the remote I? helix to R126, and finally to the 
active site (Fig. 3.12). As we mentioned previously, in class II (e.g. manganese peroxidase) and 
III peroxidases, which depend on Ca
2+
 to maintain a heme coordination environment that 
supports activity, the relevant calcium-binding site is established in part by this same portion of 
the BC loop. Therefore, it is reasonable to predict that these extended hydrogen-bonded networks 
serve in similar roles roles as those of the distal Ca
2+
 in manganese peroxidases to maintain the 
heme coordination environment. Indeed, our data are consistent with this hypothesis, disruption 
of the H-bonds between R126 and E695 by double substitutions (R126A/E695A) showed a 
greater propensity to form a hexacoordinate low-spin state as well as clear compromises to active 
site stability. 
 
3.5.3: Multiple substitutions vs. single substitution: 
The most obvious impact of multiple substitutions is that double substitutions have a 
greater effect on the thermal stability of the active site, consistent with our hypothesis that the 
remote I?-helix controls the active site integrity through these interactions. In addition, the EPR 
spectra of R126A/E695A and W159A/E695A both exhibited highly anisotropic low-spin signals 
termed HALS-type EPR spectra.  This type of spectrum, as distinct the more typical rhombic  
 130 
 
 
Figure 3.11: Interactions between R126, W159 and E695 and a water-mediated hydrogen-
bonded network along the BC loop (upper); BC loop in manganese peroxidase and Ca
2+ 
(bottom). 
Coordinates from the structure of Mycobacterium tuberculosis KatG (PDB: 2CCA) and 
manganese peroxidase were used (PDB: 1MNP). Numbering reflects E. coli KatG. 
 
 131 
 
Figure 3.12: Overlaid structures of KatGs and the water molecules. 
(PDB: 2CCA, 1SJ2 and 1MWV). Numbering reflects E. coli KatG sequence. Spheres represent 
water molecules. Red circles highlight the invariant water molecules in these three structures. 
 
 
 
 
 
 
 
 
 
 
 
 132 
signal, is purported to arise from the orientation of the planes of the imidazoles of the two 
histidine ligands.  A perpendicular orientation of the planes relative to one another produces the 
HALS-type spectrum, whereas the parallel alignment produces the typical rhombic low spin 
signal (g
x
 =2.93, g
y
 =2.28, g
z
 =1.53) (i.e., that which we observe for KatG
N
). This shift indicates 
the distal His must undergo a change in orientation relative to the heme. One possible 
explanation is multiple substitutions interrupting the water-mediated interactions between R126, 
W159 and E695 may lead to the loss of polar contacts between R126, D187 and R122.  Loss of 
these contacts along BC loop may ultimately lead to modest shifting of B-helix, allowing a shift 
in orientation of the distal His. Furthermore, the effects of multiple substitutions on active site 
heme coordination state and enzyme activity, especially for R126A/E695A, were more dramatic. 
It is noteworthy that the influence of multiple substitutions appears to be additive. This additive 
influence is also apparent in the reactivation of the stand-alone N-terminal domain. One possible 
explanation is that each one of the target amino acids plays multiple roles, one of which is 
through interaction with this connection to the active site.  
 
3.5.4: E695Q vs. E695A: 
Interestingly, substitution of E695 with either alanine or glutamine produced proteins with 
similar stability as well as spectroscopic and kinetic properties, indicating the water molecule-
mediated polar contacts between E695 and R126 are essential. In addition to association with 
W159 and R126, E695 appears to participate in another network including R691 and R144 
through a water matrix. Given that these are all polar contacts, it seems odd that E695Q would be 
more similar in properties to E695A than the wild-type enzyme.  However, given that all of the 
residues in these networks with E695 are H-bond donors (W159 indole N-H, R144, and R126), 
 133 
to accommodate the appropriate H-bonding through intervening H
2
O molecules. Therefore, a 
side chain that behaves as an exclusive H-bond acceptor is necessary. In this, glutamate but not 
glutamine, is ideal. Therefore, even the very subtle modification E695Q was sufficient to 
substantially destabilize the enzyme active site. 
 
In summary, our previous studies demonstrated a central role for the I?-helix of the C-
terminal domain in maintaining active site conformation and stability. A large distance separates 
this structure from the active site, and it was not clear how this effect is communicated to the 
active site. Available structures of KatG reveal that E695 in I?-helix is a central component of 
several hydrogen-bonded networks also including strictly conserved R126 and W159 from the 
adjacent N-terminal domain and a well-ordered water matrix from domain interface to active site 
which appear to stabilize the BC loop. Indeed, disruption of these interactions in the intact 
protein had little effect on the catalytic function and the stability of global structure of the protein 
but had a profound impact on the stability of the active site. In addition, substitutions of these 
residues substantially decreased the reactivation of KatG
N
 by KatG
C
. These studies demonstrate 
a clear connection between the I?-helix and active site conformation and function. 
 
 
 
 
 
 
 
 134 
 
 
 
 
 
CHAPTER FOUR 
 
 
BORROWING THE E.COLI CATALASE-PEROXIDASE C-TERMINAL DOMAIN AS A  
SCAFFOLD FOR GENERATION OF NEW HEME-DEPENDENT CATALYSTS 
 
4.1 Introduction:  
Of reactive oxygen species, H
2
O
2
 is the most frequently observed in biological systems. In 
most organisms, H
2
O
2
 is scavenged by either catalases or peroxidases [147]. Catalase-
peroxidases (KatGs) are unique in that they use a single active site to carry out both peroxidase 
and substantial catalase activities. KatGs all have a two-domain structure, where each domain 
resembles a plant peroxidase with respect to its sequence and three-dimensional structure. These 
data suggest that the two-domain structure arose from the fusion of two adjacent genes following 
a gene-duplication event [118, 119].  In the intervening time, it would appear that the N-terminal 
domain has maintained the heme-dependent active site. In contrast, the C-terminal domain has 
lost its ability to bind heme and catalyze any reaction of its own. Due to its large separation from 
the N-terminal domain active site (?30 ?), it cannot play a direct role in the catalytic function of 
that domain either. Nevertheless, the C-terminal domain is essential to KatG function as without 
it, the protein is no longer expressed in soluble form but is instead found in inclusion bodies. In 
addition, following denatured purification, refolding and reconstitution, the stand-alone N-
terminal domain (KatG
N
) shows a shift in its active site structure, allowing the direct 
 135 
coordination of the distal histidine to the heme iron. Consequently, KatG
N
 has neither catalase 
nor peroxidase activity [142, 144]. 
 
Sequence alignments and structural comparisons show that the ?inactive? C-terminal 
domain retains all 10 major helices of a typical peroxidase, strongly resembles the N-terminal 
domain and other members of the peroxidase-catalase superfamily (e.g. cytochrome c peroxidase 
[CCP] and ascorbate peroxidase [APX]) [119]. Indeed, an overlay of CCP and the KatG C-
terminal domain indicates that the latter retains the helical architecture and topological 
arrangement of a typical plant peroxidase [143] (Fig. 4.1).  However, even though the active site 
framework is present, the basis for the lack of catalytic function in the C-terminal domain is also 
readily apparent. With respect to heme binding, there are two side chains within the C-terminal 
domain vestigial ?active site? that are conspicuously different than the functional active sites of 
the peroxidase-catalase superfamily. First, the typical His ligand for the heme has been replaced 
by an Arg (617 by E. coli numbering). Adjacent to R617, M616 appears in place of a glycine, 
alanine or serine residue. The large aliphatic methionine side chain protrudes into the cavity 
normally occupied by the heme. The other difference is the presence of a loop that serves to 
insert an additional hydrophobic side chain into the space vacated by the heme and to cap access 
to the site using an Arg side chain  (Fig. 4.2).  
 
The typical distal side histidine, a general acid/base catalyst, is replaced by A473 [182-
184]. Likewise, at position 469, the arginine purported to stabilize developing negative charge on 
the distal oxygen of the peroxide substrate is replaced by a serine [182, 183, 185].  
 136 
 
Figure 4.1: Overlaid structures of KatG
C
 (green) and CCP (cyan) with a bound Heme 
(red).  
 
 
 
 
 
 
 
 
 
 137 
 
 
Figure 4.2: Comparison of the ?active site? environments of APX and KatG
C
. 
APX (PDB: 1OAF), KatG
C
 (PDB: 1U2K) 
 
 
 
 138 
Interestingly, the distal active site tryptophan is retained as W472, and the H-bond partners 
of the proximal and distal histidines are retained as D681 and N503, respectively (Fig. 4.2) [70, 
186-190]. Taken together, these data indicate that the C-terminal domain may provide an ideal 
?blank slate? for engineering new heme-dependent catalysts. The objective of this study was to 
capitalize on the existing active site framework of the C-terminal domain as a starting point for 
engineering new heme-dependent peroxidative catalysts based on the separately isolated C-
terminal domain (KatG
C
). Here, UV-vis absorption, MCD and EPR spectroscopy were used to 
evaluate the heme-binding environment, and steady kinetic analyses were employed to evaluate 
the restoration of peroxidase activity. 
 
4.2 Materials and methods: 
4.2.1 Materials: 
Hydrogen peroxide (30%), imidazole, hemin, ampicillin, chloramphenicol, sodium 
dithionite, phenylmethylsulfonyl fluoride (PMSF), 2,2?-azino-bis (3-ethylbenzthiazoline-6-
sulfonate) (ABTS) were purchased from Sigma (St. Louis, MO). Isopropyl-?-d-
thiogalactopyranoside (IPTG), urea, mono- and dibasic sodium phosphate, acetic acid, and 
sodium acetate were obtained from Fisher (Pittsburgh, PA). Benzonase was purchased from 
Novagen (Madison, WI). All restriction enzymes were purchased from New England Biolabs 
(Beverly, MA). All oligonucleotide primers were purchased from Invitrogen (Carlsbad, CA). All 
E. coli strains (BL-21-Gold [DE3] pLysS and XL-1 Blue), Pfu polymerase, and T4 DNA ligase 
were obtained from Agilent (La Jolla, CA). Nickel-nitrilotriacetic acid (Ni-NTA) resin was 
purchased from Qiagen (Valencia, CA). All buffers and media were prepared using water 
purified through a Barnstead EasyPure II system (18.2 M?/cm resistivity). 
 139 
4.2.2 Cloning: 
With the exception of the M616A/R617H variant, plasmids encoding variants of KatG
C
 
were prepared using mutagenic coding and non-coding primers according to the ?Round-the-
Horn? procedure [152]. The M616A/R617H variants expression construct was prepared using 
the QuickChange procedure with the complementary primers (Table 4.1). Amplification products 
from ?Round-the-Horn? procedure were subjected to blunt-end ligation with T4 DNA ligase and 
used to transform E. coli (XL-1) Blue by a standard heat shock procedure. Plasmids from 
candidate colonies were isolated based on conferred ampicillin resistance and evaluated by 
diagnostic restriction digest (see Table 4.1). Plasmids from positive candidates were subjected to 
DNA sequence analysis to verify that the correct mutation had been incorporated without any 
unintended mutations. Plasmids verified to carry the correct mutations were used to transform 
our E. coli  expression host (BL-21 Gold [DE3] pLysS).  
 
4.2.3 Expression and Puri?cation:  
Expression was carried out in E. coli (BL-21-Gold [DE3] pLysS) using cultures 
supplemented with ampicillin and chloramphenicol as previously described. The KatG
C
 was 
expressed in a soluble form and purified as previously described [144]. KatG
N
, and all variants of 
KatG
C
 variants were expressed in inclusion bodies. Therefore, purification was carried out 
according to the procedure previously described for KatG
N
 (see Chapter 2). Concentrations of 
purified enzyme were estimated according to the method of Gill and von Hippel [154]. 
 
 
 
 140 
 
 
Table 4.1: Primer sets for site-directed mutagenesis of KatG
C
 
 
 Primer Sequences: 
Restriction 
enzymes 
M616G/R617H 
TGTGGCGCGCACGTACTGGTG 
CACCAGTACGTGCGCGCCACA 
 
Sph I 
G614A/M616G/R617H 
5?-Phos-GCGCGTCATGTACTGGTG-3? 
5?-Phos-CACAGCGCAGTCATTCGC-3? 
 
Ngo MI 
?575-582 
5?-Phos-CGCGCTCGTCTGACGTT-3? 
5?-Phos-TGCTCAGCAGCTCAACA-3? 
 
Mfe I 
A473H 
5?-Phos-TGCTGCACAGCGCTCTACT-3? 
5?-Phos-CGATACAGCTCACTACAGACAGACAGATC-3? 
 
Afe I 
 S469R/A473H 
5?-Phos-CGTGCTGCATCTGCTCTACTC-3? 
5?-Phos-CGTACAGCTCACTACAGACAGACAGATCG-3? 
Mlu I 
 
 
 
 
 
 
 
 
 
 
 141 
4.2.4 Enzyme reconstitution and heme absorption spectra: 
Due to the poor solubility of hemin at neutral and acidic pH, hemin solutions for 
reconstitution were prepared by dissolving solid hemin in 0.1 M KOH. Hemin concentration was 
determined by the method of Falk [155] and 0.75 equivalents added to modified KatG
C
 variants. 
These solutions were allowed to incubate for at least 24 h at 4 ?C. The reconstituted enzyme 
solution was then centrifuged in order to remove precipitated heme and other insoluble material. 
Concentrations of reconstituted enzymes were determined using the pyridine hemichrome assay 
[155]. Protein containing heme in the ferrous state was prepared by adding a small amount (<10 
mg) of sodium dithionite to the ferric form of the enzyme. All spectra were obtained at room 
temperature on a Shimadzu UV-1601 spectrophotometer (Columbia, MD) with a cell path length 
of 1.0 cm. 
 
4.2.5 Peroxidase activity assays and spectroscopy: 
Peroxidase assays and initial velocity fitting were performed as described in Chapter 2. If 
inhibition was evident, the fitting equation was modified to a general substrate-dependent 
inhibition model (Chapter 2) [191]. Settings and parameters of (magnetic) circular dichroism 
(CD) and Electron Paramagnetic Resonance (EPR) measurements were the same as described in 
Chapter 2.   
 
4.3 Results: 
4.3.1 Engineering strategy:  
In order to restore heme binding capability to KatG
C
, we used site-directed mutagenesis to 
substitute R617 with histidine and M616 with glycine  (HB1) (Table 4.2).  In doing so, an 
 142 
appropriate ligand for heme binding was introduced and a large aliphatic side chain obstructing 
the active site cavity was removed, respectively. This modification resulted in a contiguous 
stretch of three glycine residues (G614, G615 and M616G). We were concerned that the 
conformational flexibility resulting from three consecutive glycines would be overly disruptive 
to the KatG
C
 structure.  Thus, we replaced the first glycine in the sequence with alanine (i.e., 
G614A), a variant we designated as HB2. We anticipated that the apex of an extended loop may 
obstruct heme binding by the HB1 and HB2 vaiants. Variants from which this structure was 
excised were generated from HB1 and HB2, generating HB1a and HB2a, respectively). Of these 
variants, HB2a showed the spectroscopic features that seemed most appropriate for further 
modification. Therefore, HB2a was modified sequentially by site directed mutagenesis to 
produce the A473H and S469R/A473H variants (His as a general base/acid and Arg as an 
electrostatic catalyst) generating PX1 and PX2, respectively, with the purpose of restoring 
peroxidase activity. 
 
4.3.2 Mutagenesis, expression and purification of KatG
C
 and its variants:  
As confirmed by DNA sequence analysis, we successfully produced constructs for 
expression of all of these KatG
C
 variants. All of them were successfully expressed and isolated. 
The far-UV CD spectra recorded for each variant were indistinguishable from that of the 
unmodified KatG
C
, suggesting that the modifications did not introduce catastrophic disruptions 
of protein structure (Fig. 4.3). 
 
 
 
 143 
Table 4.2: Variants name corresponding to the detailed mutagenesis strategy 
Variants name:   
HB1 M616G/R617H KatG
C
 
HB2 G614A/M616G/R617H KatG
C
 
HB1a M616G/R617H ?575-582 KatG
C
 
HB2a G614A/M616G/R617H ?575-582 KatG
C
 
PX1 A473H/G614A/M616G/R617H ?575-582 KatG
C
 
PX2 S469R/A473H/G614A/M616G/R617H ?575-582 KatG
C
 
 144 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 4.3: Far-UV circular dichroism of KatG
C
 and KatG
C
 variants.  
Spectra were recorded at 23 ?C in 5 mM phosphate buffer, pH 7.0 using a 0.5 mm quartz cell. 
 
 
 
 
 
 
 
 
 145 
 
4.3.3 UV-vis spectroscopic properties of modified KatG
C
: 
As isolated, the stand-alone C-terminal domain (KatG
C
) does not contain heme [143].  
Consistent with this property, the spectrum of KatG
C
 incubated with 0.75 eq. hemin showed 
typical features of free hemin with an absorption maximum near 380 nm. In contrast, the UV-
visible absorption spectrum of HB1 showed a signature low-spin heme spectrum with a Soret 
band maximum near 415 nm and ? and ? band maxima at 538 nm and 558 nm, respectively. 
These features closely resembled those of the stand-alone KatG N-terminal domain of KatG 
(KatG
N
). It indicated that heme binding was restored as a result of introducing proximal His 
ligand and that the bound heme was in a hexacoordinate, low-spin ferric state. However, HB1 
showed a lower molar absorptivity than KatG
N
 (Fig. 4.4 and Table 4.3). As described previously, 
construction of the HB1 variant resulted in three consecutive Gly residues.  The greater 
conformational flexibility imparted by glycine residues is potentially problematic with respect to 
protein stability as it introduces a greater entropic tendency toward misfolded states. The variant 
we generated to minimize this problem, HB2, produced absorption spectra with similar maxima 
as those observed for HB1 but with enhanced molar absorptivity. The loop excision variants 
(HB1a and HB2a) showed a similar trend in spectra as that observed from their parent proteins 
HB1 and HB2 (Fig. 4.5). Finally, spectra recorded for the PX1 and PX2 variants were also 
dominated by hexacoordinate low-spin features. Of all the variants, overall, PX1 and PX2 
spectra were most similar to those observed for KatG
N
. 
 
4.3.4 MCD spectroscopic properties of modified KatG
C
: 
In order to more effectively differentiate the high-spin and low-spin state contributions of  
 146 
 
300 350 400 450 475 525 575 625 675
0
20
40
60
80
100
120
0.0
2.5
5.0
7.5
10.0
12.5
15.0
Ferric HB1
KatG
C
+Heme Ferric KatG
N
 
536
568
416
414.5
532
563.5
601.5
385
Wavelength (nm)
!
 
(
m
M
-
1
 
c
m
-
1
)
!
 
(
m
M
-
1
 
c
m
-
1
)
 
Figure 4.4:  UV-vis spectra of ferric KatG
C
, KatG
N 
and KatG
C
 variants (HB 1). 
 
 
 
 
 
 
 
 
 
 
 147 
350 375 400 425 450 475 500 550 600 650
HB2
HB1a
HB2a
HB1
PX1
414.5
413.5
414.5
414
414
414.5
532  561
533  559
532  561
530.5  561.5
532  564
532  563.5
PX2
Wavelength (nm)
 
Figure 4.5:  UV-vis spectra of ferric KatG
C
, KatG
N 
and KatG
C
 variants. 
 
 
 
 
 
 
 
 
 
 
 
 
 148 
 
Table 4.3: Absorption characteristics of KatG
N
 and KatG
C
 variants 
Enzyme Absorption maxima [nm] (absorptivity [mM
-1
 cm
-1
]) 
Ferric Soret ? ? 
KatG
N
 416 (104.3) 536 (10.1) 568 (7.7) 
HB 1 414.5 (77.8) 532(8.5) 563.5 (7.0) 
HB 2 413.5 (98.9) 532 (11.7) 564 (9.8) 
HB 1a 414 (81.6) 530.5 (9.4) 561.5 (8.2) 
HB 2a 414.5 (91.5) 532 (10.1) 561 (8.8) 
PX 1 414 (103.9) 532 (10.1) 561 (8.9) 
PX 2 414.5 (101.8) 533 (9.6) 559 (8.5) 
 149 
these variants, magnetic circular dichroism (MCD) spectra were recorded and analyzed. With the 
exception of HB1 and HB2 the ferric state of all the variants showed an intense derivative feature 
centered at 415 nm (Fig. 4.6B).  This is typical of the hexacoordinate low-spin state.  One 
interpretation for the less intense spectra for HB1 and HB2 would be that there is a much greater 
contribution from high spin species. MCD spectra recorded for the ferrous states of all the 
variants (Fig. 4.6A) consistently showed an intense A-term centered around 557 nm, indicative 
of substantial contribution from hexacoordinate low-spin heme species. In the ferrous state, high-
spin heme produces an intense B-term with a maximum around 440 nm. This feature was 
particularly prominent for the HB2a and PX1 variants and less so for HB1 and HB2, indicating 
that the former pair of proteins were more favorable to high-spin heme states.  Thus, while HB2a 
and PX1 do show a greater tendency toward heme ligation environments that give rise to high-
spin heme iron, HB1 and HB2 do not.  It would seem, then, in conjunction with UV-visible and 
ferrous MCD spectra, that the less intense features of the ferric MCD spectra of HB1 and HB2 
are due to an inability of these variants to bind heme in a uniform environment. A greater 
proportion of adventitiously bound heme would be expected to contribute to broader and lower 
intensity spectra. 
 
4.3.5 EPR spectroscopic properties of modified KatG
C
: 
A more definitive assessment of the heme-binding environment was carried out using EPR 
spectroscopy (Fig. 4.7). Consistent with the low molar ellipticity in their ferric MCD spectra, 
HB1 and HB2 showed only weak signals in the low-temperature EPR spectra. However, removal 
of the ?active site? placeholder loop (i.e, HB1a and HB2a) produced a domain (and derivatives 
thereof) with more sharply defined EPR spectra. Similar to those observed for the stand-alone 
 150 
 
400 450 500 550 600 650
-100
-50
0
50
100
HB2
HB1
HB2a
PX1
PX2
Ferrous
KatG
N
HB1a
Wavelength (nm)
!
"
/
H
 
(
c
m
-
1
 
M
-
1
 
T
-
1
)
350 400 450 500 550 600 650
-100
-50
0
50
100
Ferric
Wavelength (nm)
!
"
/
H
 
(
c
m
-
1
 
M
-
1
 
T
-
1
)
A
B
 
Figure 4.6: Magnetic circular dichroism (MCD) spectra for KatG
C
 and KatG
N
 variants. 
 
 
 151 
1000 2000 3000 4000
HB2a
HB1a
HB2
HB1
KatG 
N 
!FG!DE200-214
 
PX2
PX1
g=5.99
g=2.92 g=2.28 g=1.99
g=1.95
Field  [G ]
g=1.53
 
Figure 4.7: EPR spectra for ferric state of KatG
C
 and KatG
N
 variants. 
 
 
 
 
 
 
 
 
 
 152 
N-terminal, the most prominent features were observed near g=2.92 and g=2.28, consistent with 
a hexacoordinate low-spin Fe (III) complex. Notably, for PX1, a much more pronounced feature 
was detected at g=5.99, indicating the presence of a greater proportion of a high-spin state. 
However, the EPR spectra of PX2, where Arg had been introduced as an electrostatic catalyst 
showed a diminished intensity in high-spin state along with increased intensity of signals 
corresponding to low-spin species (Fig. 4.7). 
 
4.4.6 Steady-state kinetics of KatG
C
:  
To date, none of the modifications to KatG
C
 have produced an enzyme with catalase and 
peroxidase activities.  
 
4.4 Discussion and conclusion:  
Due to its sequential and structural resemblance to monofunctional peroxidases, the C-
terminal domain retains the general architecture of a functional peroxidase active site. Indeed, 
the sequence alignments and the structures solved for KatG?s indicate that the C-terminal domain 
retains an active site helical topology essentially identical to that observed in the N-terminal 
domain of KatG as well as other class I peroxidases. However, we speculate that side chain 
substitutions and other modifications accumulated since gene duplication and fusion have 
rendered the site unable to bind heme or have catalytic activity of its own.  
 
We surmised that this set of circumstances presented the C-terminal domain as an ideal 
?blank slate? for generating new heme-dependent catalysts. Strategically, it seemed that a 
reintroduction of peroxidatic activity would be the most straightforward approach to evaluate this 
 153 
hypothesis. Consistent with our hypothesis, dramatic improvement in heme binding was 
observed with the double substitution variant HB1 (i.e., M616G/R617H) where a steric 
impediment to occupancy of the active site by heme was eliminated and an appropriate ligand to 
the heme iron was introduced, respectively. However, the preponderance of spectral evidence 
(reduced molar absorptivity and reduced molar ellipticity in UV-vis and ferric MCD spectra, 
respectively, and the absence of clearly defined EPR spectral features) indicated a large diversity 
of heme association states. The data suggested a need for further modification to produce more 
uniformly bound heme. 
 
Visual inspection of a crystal structure solved for the E. coli KatG C-terminal domain 
shows F579 and R580 located at the apex of a loop structure oriented toward the former active 
site [143].  In this, the structure is reminiscent of Large Loop 1 (LL1) found in the N-terminal 
domain. The phenylalanine at position 579 appears to be a hydrophobic placeholder for the now 
absent heme and the arginine at position 580 obstructs access to the active site.  Both of these 
residues and the loop which bears them would be expected to limit appropriate heme binding in 
HB1 and HB2. Consistent with this hypothesis, excision of the loop (i.e., ?575-582) in 
combination with M616G/R617H substitution produced proteins with more sharply defined 
spectral characteristics. In particular with HB2a, the EPR spectrum suggested a successful mimic 
of the stand-alone N-terminal domain. 
 
For the native class I plant peroxidases, the distal His-Arg-Trp triad is strongly conserved 
and essential for peroxidative catalysis, especially for compound I formation [182-184, 25]. The 
distal histidine is employed as a general base/acid during catalysis [182-184]. As expected, 
 154 
substitution of histidine with alanine substantially decreases the rate of compound I formation in 
cytochrome c peroxidase (CCP) [192]. The presence of Ala in the analogous position in KatG
C
, 
HB1, HB2, HB1a, and HB2a would account for a lack of peroxidase activity in those proteins. In 
an attempt to introduce peroxidase catalytic ability, we replaced A473 with histidine in the HB2a 
variant to generate the PX1 protein; the protein with heme characteristics most similar to KatG
N
. 
However, no apparent peroxidase activity was detected in PX1. In order to restore the 
peroxidatic activity, an arginine was introduced distal to the heme group by replacing S469 (i.e., 
PX2). Crystal structures of CCP have revealed that the distal arginine has ?in? and ?out? 
conformations, and only the ?in? position allows the arginine side chain to form an H-bond with 
iron-linked oxygen [99]. Therefore, the arginine is proposed to stabilize the ferryl-oxo center in 
of compound I and presumably compound II.  
 
However, these modifications do not appear to be sufficient to introduce peroxidatic 
activity, possibly attributed to both ?active sites? in PX1 and PX2 having been shifted, allowing 
distal residues to directly coordinate the heme iron. As previously described (see Chapters 2 and 
3), KatG without its C-terminal domain (KatG
N
) is no longer expressed in a soluble form but is 
instead found in inclusion bodies. In addition, following denatured purification, refolding and 
reconstitution, KatG
N
 shows a shifted active site structure, allowing what appears to be the direct 
coordination of the heme iron by the distal histidine [142]. This can be remedied by introducing 
the stand-alone C-terminal domain [144]. Interestingly, the stand-alone C-terminal domain is 
expressed in a soluble form and demonstrates substantial structural stability on its own.  
However, variants directed toward reintroduction of heme binding and catalytic activity were no 
longer expressed in a soluble form. Moreover, after denatured purification, refolding and 
 155 
reconstitution, the ?active site? heme environment appears to be highly similar to that of KatG
N
, 
indicating our modifications successfully produced a mimic of KatG
N 
where similar structural 
elements (i.e, an extra domain such as C-terminal domain in KatG or a helical structure such as 
I?-helix in KatG) may be necessary to stabilize the active site conformation and thereby modulate 
the catalytic ability. Therefore, future modification will focus on finding certain structural 
elements to direct the active site return of high-spin species, allowing the distal His to exert its 
function as a general base/acid [182-184].  
 
In addition, the distal side arginine is critical and must be appropriately positioned to allow 
for the formation of compound I upon reaction with H
2
O
2
. In addition, Asp 681 and Asn 503 
fulfill the function of active site His H-bond partners. However, our substitutions may not be 
ideally situated to reestablish these critical hydrogen bonded interactions. Therefore, additional 
modifications will also focus on reorienting and adjusting residues to optimize active site H-
bonding to fully restore peroxidase activity. 
 
Due to the sequence and structural similarity between the C-terminal domain of KatG and 
typical plant peroxidases, the most straightforward approach of engineering KatG
C
 is 
reintroduction of peroxidatic activity by restructuring a typical plant peroxidase active site. In 
addition, KatG
C
 also provides a practical starting point for engineering other heme-dependent 
enzymes. First, the existing structure and its ?active site? framework could be exploited to 
engineer other types of peroxidases. For instance, a DyP-type peroxidase active site may be 
obtainable by introducing aspartate as an acid/base catalyst starting with the HB2a template. 
Likewise, Cys instead of His, can be introduced as the proximal heme-binding ligand; in the 
 156 
distal site, Glu rather than His is used as catalytic acid/base to engineer KatG
C
 into a fungal 
heme-dependent chloroperoxidase (see Chapter 1). Second, the KatG active site is highly 
versatile: catalase, peroxidase, dioxygenase, and peroxynitritase activities have all been observed 
[193, 194, 195]. The existing KatG
C
 active site framework can be refined toward a particular 
reaction and set of substrates rather than building an entirely new activity from first principles.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 157 
 
 
 
 
 
CHAPTER FIVE 
 
 
RAPID DIAGNOSIS OF MET-TYR-TRP COVALENT ADDUCT FORMATION IN 
KATG BY SDS-PAGE 
 
5.1 Introduction: 
Catalase-peroxidases (KatGs) are bifunctional enzymes, using a single active site to 
degrade H
2
O
2
 by either a catalase or peroxidase mechanism. All available crystal structures for 
intact KatG show two unique covalent bonds between three amino acid side chains, W105, Y226, 
and M252 (E. coli numbering). This adduct is located on the distal side of the heme (Fig. 5.1) 
[119-121, 131], where it appears to be indispensable for the unique catalase activity of KatG 
[132, 134, 135, 149, 196-198]. The formation of this covalent adduct is proposed to be 
autocatalytic wherein a peroxide oxidizes the KatG heme to a compound I (i.e. Fe
IV
=O 
[porphyrin
.+
]) species, which then oxidizes the Trp and Tyr residues leading to establishment of 
the adduct by radical-radical coupling [135]. 
 
In addition to crystal structures for various KatGs, the M-Y-W adduct has also been 
observed in the enzyme in solution by LC/MS/MS peptide mapping techniques [132-135]. 
Evaluation of this adduct by mass spectrometric approaches requires a relatively large amount of 
protein, long analysis time, and expensive instrumentation.  
 
 
 158 
 
 
Figure 5.1: Distal active site M-Y-W covalent adduct in KatG from M. tuberculosis  
(PDB: 1SJ2).  
 
 
 
 
 
 
 
 
 159 
It has been reported that SDS-PAGE separation of recombinant KatP (a periplasmic 
catalase?peroxidase from Escherichia coli O157:H7) resulted in two forms of the protein [153]. 
Each band was determined to be a single protein, and each one was identified as KatP by HPLC 
as well as peptide sequence analysis. The presence of heme and catalase activity was only 
observed in the protein with lower apparent molecular weight [153]. Similarly, two forms of 
protein have been observed by SDS-PAGE in our laboratory with our reconstituted wild-type 
KatG (E.coli K-12). A reasonable explanation for these observations is that these post-
translational covalent bonds are established within a subunit and not subject to disruption by 
SDS or thiol-based reductants. In theory then, KatG bearing the covalent adduct should be 
prohibited from completely denaturing in the presence of SDS, and as a result, should have a 
lower hydrodynamic volume, and more rapid migration through the acrylamide matrix than 
KatG lacking this novel covalent structure. 
 
5.2 Methods: 
Expression and purification of wtKatG and ?LL2 variants [151] as well as KatG
C
 and 
KatG
N
 were carried out as previously described [142, 144]. Likewise, expression and purification 
of ?200-214, ?209-228 and Y226F followed procedures that we have recently described [136]. 
The reconstituted protein solutions were allowed to incubate for at least 24 h at 4 ?C.  These 
solutions were then centrifuged to remove insoluble unincorporated heme and other debris. 
Solutions containing 1:1 ratios of KatG
N
 and KatG
C
 were incubated for 24 hours in the presence 
of 50 mM phosphate buffer, pH 7.0, 50 mM NaCl at 4 ?C as previously described in Chapter 2 
and 3.  
 
 160 
For analyses by SDS-PAGE, a 7.6% acrylamide resolving gel and 4% acrylamide stacking 
gel were used. Protein samples (0.015 nmol) were loaded and separated. In order to avoid 
peracetic acid (PAA)-induced protein aggregation, protein samples were mixed with SDS 
loading buffer immediately after treatment with 2 equivalents of PAA.  
 
5.3 Results: 
It has been shown that the M-Y-W covalent adduct is essential for the catalase activity of 
KatG, and its formation in KatG is heme-dependent. E. coli KatG expressed in a T7-based 
system produces the enzyme predominantly (~95%) in its apoprotein form [151]. Apo-KatG can 
be reconstituted with heme to produce an enzyme with full catalase and peroxidase activities 
[151]. As such, this establishes a system by which to evaluate covalent adduct formation by 
SDS-PAGE. Without added heme, a single band (~80 kD) is observed by SDS-PAGE (Fig. 
5.2A). Reconstitution of KatG with hemin resulted in the appearance of a second band with a 
lower apparent molecular weight, and the relative intensity of this band increased with an 
increasing amount of heme added up to 1 molar equivalent. Above 1 equivalent, no further 
increases in the intensity of the lower band were observed [199] (Fig. 5.2A). 
  
Reconstitution of KatG with redox-inactive Zn-protoporphyrin IX prevented the 
appearance of the lower band as one would anticipate given the oxidative chemistry necessary to 
establish the M-Y-W adduct.  It has been shown previously that full establishment of the adduct 
requires hydroperoxide.  Accordingly, hemin-reconstituted KatG treated with two equivalents of 
PAA migrated almost entirely as the lower apparent molecular weight band.  Conversely, KatG 
reconstituted with Zn-protoporphyrin IX and then reacted with 2 equivalents of PAA showed no  
 161 
M 1 2 3 4 5 6
A
B
C
75 KD
75 KD
75 KD
 
Figure 5.2:  Evaluation of M-Y-W covalent adduct formation by SDS-PAGE [199]. 
(A) Denaturing electrophoretic separation of apo-KatG (lane 1) was reconstituted with 0.3 
equivalents (lane 2), 0.6 equivalents (lane 3), 1.0 equivalent (lane 4), 2.0 equivalents (lane 5) and 
5.0 equivalents (lane 6) of heme. Marker (M) was shown as 75 KD. (B) Effect of heme and 
peroxide on migration of apo-KatG. Apo-KatG (lane 1) with heme (lanes 2 and 5), Zn-PPIX 
(lanes 3 and 6), and 2 equivalents of PAA (lanes 4-6) added. (C) Effect of substituting Y226 on 
the protein migration. Apo-Y226F (lane 1) with heme (lanes 2 and 5), Zn-PPIX (lanes 3 and 6), 
and 2 equivalents of PAA (lanes 4-6) added. 
 
 
 
 
 
 
 
 
 162 
shift in SDS-PAGE migration compared to the untreated Zn-protein or apo-KatG [199] (Fig. 
5.2B).  Substitution of the critical tyrosine (Y226 in E. coli KatG) with phenylalanine prevents 
formation of the covalent adduct and eliminates catalase activity [128, 136]. With Y226F KatG 
only the higher apparent molecular weight band was observed regardless of reconstitution (with 
hemin or Zn-protoporphyrin IX) or treatment with PAA [199] (Fig. 5.3C).  
 
We have applied this SDS-PAGE based technique to other KatG variants designed to target 
various aspects of structure and function. For example, the strictly conserved Tyr residue (Y226) 
is contained within the KatG-unique structure Large Loop 1 (LL1). In order to evaluate the role 
of the larger LL1 structure in KatG function, we have generated two deletion variants (?200-214 
and ?209-228) [136]; the latter (?209-228) lacks Y226, while the former does not.   
 
Analyses by SDS-PAGE gel showed the presence of two bands for ?200-214, but only one 
for ? 209-228.  Accounting for the difference in molecular weight caused by the deletion, ?200-
214 showed an increased proportion of protein migrating as the lower molecular weight band 
following treatment with two equivalents of PAA, consistent with the wild-type enzyme. 
Accounting for the decreased molecular weight caused by the deletion, ?209-228 showed 
behavior analogous to that of Y226F KatG  (Fig. 5.3A). 
 
Large Loop 2 (LL 2) is another structure unique to KatG.  It has been shown to be essential 
for the catalase activity of KatGs.  Indeed, deletion of LL2 or portions thereof produces  proteins 
with a 99.8% decrease in catalase activity [151, 130]. In these variants, the mechanistic 
explanation for the loss of catalase activity is less obvious.  The LL2 structure does not contain  
 163 
M1 2 3 4 5 6
A
M1 2 3 4 5 6
B
M
C
1 2 3 4 5 6
75 KD
75 KD
50 KD
25 KD
 
Figure 5.3:  The influence of peripheral structures on the formation of M-Y-W adduct. 
(A) SDS-PAGE separation of ?209-228 and ?200-214 prior to reconstitution with heme (lane 1 
and 4, respectively) following reconstitution with 1 equivalent of heme (lanes 2 and 5, 
respectively), and treatment of reconstituted ?209-228 and ?200-214 with 2 equivalents of PAA 
(lanes 3 and 6, respectively). (B) Effect of deletion of LL2 on formation of covalent adduct. 
Apo-wtKatG and apo-?LL2 (lanes 1 and 4, respectively) following reconstitution with heme 
(lanes 2 and 5, respectively) and treatment with 2 equivalents of PAA (lanes 3 and 6, 
respectively). (C) Effect of removal and reintroduction of the C-terminal domain on the 
formation of MYW covalent adduct. Lanes 1-3 showed KatG
N
 reconstituted with heme (lane 1), 
in presence of 200 equivalents of H
2
O
2
 (lane 2) or in presence of 2 equivalents of PAA (lane 3). 
Lanes 4-6 showed the domain mixture (KatG
N
+KatG
C
) reconstituted with heme (lane 6), in 
presence of 200 equivalents of H
2
O
2
 (lane 5) or in presence of 2 equivalents of PAA (lane 4). 
 
 164 
any of the amino acids of the M-Y-W adduct. As such, the loss of catalase activity might be due 
to a failure to establish this adduct or due to some other mechanisms. Here, a straightforward 
analysis of covalent adduct formation by SDS-PAGE helps to deconvolute the role(s) of a 
structure like LL2.  We took the KatG deletion variant lacking the entire LL2 structure (KatG 
[?LL2]).  Reconstitution of this protein with 1 equivalent of heme and subsequent exposure of 
reconstituted enzyme to 2 equivalents of PAA produced a similar SDS-PAGE migration shift as 
those observed for wild-type KatG (Fig. 5.3B). Independent evaluation of covalent adduct 
formation by MS techniques verified the presence of the covalent adduct by peptide mapping in 
KatG (?LL2) [130]. 
 
As described elsewhere, the C-terminal domain of KatG likely originated from a gene 
duplication and fusion event, and this structure is located 30 ? away from the active site [118].  
From this position it appears to serve as a platform to direct active site structural adjustment (see 
Chapters 2 and 3). Deletion of the C-terminal domain produces a protein (KatG
N
) in which the 
structure of the active site has shifted, allowing what appears to be the direct coordination of the 
heme iron by the distal histidine [142]. 
 
The expression of KatG
N
 in inclusion bodies, its purification under denaturing conditions, 
reconstitution with hemin following purification, and lack of catalase activity following 
reconstitution would suggest that the covalent adduct is not formed in this protein.  However, 
KatG
N
 incubation with the stand-alone C-terminal domain (KatG
C
) does result in the 
reemergence of catalase activity, suggesting that the covalent adduct is formed at least to some 
extent when KatG
N
 and KatG
C
 are incubated together. Here again, SDS-PAGE provides a 
 165 
straightforward way to qualitatively assess this hypothesis.  Analyses of KatG
N
 migration by 
SDS-PAGE showed a single band for the protein even following reconstitution with hemin and 
exposure to PAA, indicating that without the C-terminal domain, KatG
N
 was incapable of 
establishing the M-Y-W adduct (Fig. 5.3C). Interestingly, incubation of KatG
N
 with 200 molar 
equivalents of H
2
O
2
 resulted in the loss of protein at the expected molecular weight.  Due to its 
inability to consume H
2
O
2 
by either catalatic or peroxidatic turnover, excess H
2
O
2
 appears to 
result in extensive oxidation of the protein as well as subsequent aggregation (Fig. 5.3C).  In 
contrast, our SDS-PAGE analyses indicated that incubation of KatG
N
 with equimolar KatG
C
 did 
allow for the formation of M-Y-W adduct (Fig. 5.3C). Furthermore, the introduction of KatG
C
 
clearly enhanced the ability of KatG
N
 to metabolize large quantities of H
2
O
2
, consistent with an 
enzyme that has substantial catalase activity. 
 
5.4 Conclusions: 
SDS-PAGE is universally applied in biochemical research to separate proteins based upon 
their subunit molecular weight. Interestingly, as all residues contributing to the M-Y-W adduct 
are found within the same subunit, KatG bearing the covalent adduct should 1) be prohibited 
from completely denaturing in the presence of SDS, 2) have a lower hydrodynamic volume, and 
therefore, 3) migrate more rapidly through the acrylamide matrix than KatG lacking this novel 
covalent structure. For KatGs, this migration shift was easy to identify using SDS-PAGE. 
Compared to LC/MS/MS approaches, SDS-PAGE is more rapid and easily executed. Less 
sample is required, and it is less time and material intensive to quickly and qualitatively evaluate 
the influence of a range of conditions on formation of the adduct.  Therefore, it can be used as a 
complementary procedure to more intensive LC/MS/MS procedures. By this fast diagnosis, we 
 166 
observed the formation of this adduct in KatG required a redox active protoporphyrin and was 
aided by the addition of an oxidant like peracetic acid. In addition, we demonstrated that the 
technique was useful for evaluating variants like KatG (?LL2) where the mechanistic 
explanation for the loss of catalase activity is unclear.  Likewise, the technique is useful for 
making a determination regarding the timing for M-Y-W adduct formation in a protein like 
KatG
N
 which only gains catalatic activity in the presence of a partner protein, KatG
C
.  In 
conclusion, we suggest that this technique is not only useful for evaluating KatG, but also for 
other proteins where the establishment of novel post-translational covalent adducts leads to 
enzyme activation or inactivation. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 167 
 
 
 
 
 
CHAPTER SIX 
 
 
SUMMARY 
 
 
Catalases and peroxidases scavenge H
2
O
2
, the most abundant of the reactive oxygen 
species. By direct and indirect mechanisms, H
2
O
2
 is able to inflict damage on most cellular 
components, and it is generated as an inevitable consequence of aerobic metabolism.  As a result, 
catalases and peroxidases are essential protective enzymes found in all aerobic organisms.  
Interestingly, copious production of H
2
O
2
 forms a critical component of the inflammatory 
responses of higher eukaryotes against infection by pathogenic organisms. For these pathogens, 
defenses against the accumulation of H
2
O
2
 are particularly important for their survival within the 
host organism.  
 
Heme-dependent catalases are able to rapidly and efficiently disproportionate H
2
O
2
, 
generating H
2
O and O
2
 as relatively harmless products.  These enzymes do so with impressive 
second order rate constants ranging from 10
6
 to 10
7
 M
-1
s
-1
. In contrast to the catalases, heme-
dependent peroxidases degrade H
2
O
2
 at the expense of exogenous reducing substrates (i.e., 
electron donors) whose oxidation products often have other important physiological functions. 
Therefore, peroxidases have a variety of essential but organism/cell-specific physiological roles 
in addition to reducing oxidative stress. Catalase-peroxidase (KatG) are unique in nature in that 
they are capable of degrading H
2
O
2 
via peroxidase and catalase activity, both of which are at 
levels comparable to those monofunctional counterparts. KatG is found exclusively in 
 168 
prokaryotic and lower eukaryotic organisms where it serves as an essential component of the 
oxidative stress response.  
 
The structure of a heme-dependent protein and its connection to functionality can be 
evaluated at four levels: 1) the structure of the heme prosthetic group itself, 2) the heme 
environment as dictated by protein-derived ligand(s), 3) the immediate heme environment as 
dictated by protein-derived groups other than the ligands, and 4) protein features peripheral to, 
and in some cases, distant from the active site. The research described in this dissertation tackles 
the structure-function equation on all of these levels. Certainly, the most convoluted and most 
difficult to address is level four. KatGs have a novel two-domain structure. Neither domain bears 
any resemblance to monofunctional catalases but both domains strongly resemble the overall 
fold monofunctional peroxidases. In fact, the N-terminal domain has an active site that is 
virtually indistinguishable from those of monofunctional peroxidases. Given the high similarity 
of KatG and monofunctional peroxidase in terms of heme structure, iron coordination and 
protein-derived heme environment, it is reasonable to suggest the protein structures/features 
external to the KatG active site are essential for imparting the hallmark bifunctional catalytic 
abilities of the enzyme. Therefore, KatG provides an excellent opening to evaluate the 
contributions of distant protein structure to active site structure and function.  
 
Comparisons of the protein structures and amino acid sequences of KatGs versus 
monofunctional peroxidases reveal three structural features unique to KatG: two interhelical 
insertions (LL1 and LL2) and a C-terminal domain. The Goodwin laboratory has shown that all 
of these structures have a profound influence on the function of KatG, but the mechanism by 
 169 
which this distant control is exerted is still poorly understood, particularly, the C-terminal 
domain. The crystal structure of the stand-alone C-terminal domain reveals that the features of 
easy folding, protease-resistance, the conserved residues pattern involved in inter- and intra- 
subunit interactions make the C-terminal domain has the appearance of a platform to direct the 
proper folding of the N-terminal domain.  
 
1: The integral role of I?-helix in the function of C-terminal domain: 
I?-helix appears at the interface between the two domains. It makes contacts with several 
elements from the N-terminal domain of the adjacent subunit using strongly conserved Leu 690, 
Arg 691, Glu 695, and Tyr 697. As such, it has all the appearances of a platform by which the C-
terminal domain may direct structural adjustments to the N-terminal domain and modulate active 
site conformation. Our data showed multiple substitutions of I?-helix substantially increased the 
propensity of KatG to form the hexacoordinate low-spin state as well as clear compromises to 
active site and overall protein structural stability. Indeed, the quadruple variant showed 
properties nearly identical to KatG
N
 in all respects even though it retained what appeared to be a 
correctly folded C-terminal domain. 
 
Another important outcome of the work is that it showed the separated domains as an 
advantageous tool to address these distant interactions. We have capitalized on the unique ability 
of KatG
C
 to reactivate an otherwise inactive KatG
N
 in order to show that the C-terminal domain's 
I?-helix is a central component of a platform to facilitate folding events and/or conformational 
adjustments in the N-terminal domain. Indeed, our data indicate that even single substitutions to 
the I?-helix substantially limit the ability of KatG
C
 in this respect. With two substitutions to the 
 170 
I?-helix, KatG
C
 became completely incapable of directing the restructuring of KatG
N
 for 
reactivation. Taken together, we demonstrated that the C-terminal domain I? helix is essential to 
direct structural adjustments in the adjacent N-terminal domain. 
 
2: Distant interactions connecting the remote I?-helix to active site conformation, stability and 
catalytic function: 
Though we established the I?-helix as an important structure for conformational control of 
the active site, how this effect is communicated to the active site remaines to be determined.  We 
noted some important interactions with E695, a residue that seemed particularly important to the 
function of the I?-helix. Using a similar approach we observed that interactions of E695 with two 
other residues, R126 and W159 were also important. Most strikingly, we found that disruption of 
these interactions in the intact protein had little effect on the stability of the global structure of 
the protein but a profound impact on the stability of the active site.  Even the subtle modification 
represented in the E695Q substitution was sufficient to substantially destabilize the enzyme 
active site.  Here again, the behavior of the separately expressed and isolated domains indicated 
that these residues are essential for C-terminal domain-dependent control of the N-terminal 
domain active site. 
 
The primary significance of this work is that it reveals a structural basis for a connection 
via remote interactions to active site conformation. Well-ordered water molecules from the 
domain interface to the active site appear to stabilize the BC loop, which further adjust the B-
helix bearing distal His. The importance of this structure is apparent from the attention it receives 
in terms of increasing its stability in enzymes related to KatG. In Class II (e.g. manganese 
 171 
peroxidase) and III peroxidases, which depend on Ca
2+
 to maintain the heme coordination 
environment that supports activity, the relevant calcium-binding site is established in part by this 
same portion of the BC loop. The roles of these polar contacts may resemble that of Ca
2+
 in class 
II peroxidases, providing a possible explanation for the evolutionary persistence of the C-
terminal domain dedicated to stabilizing the catalytic active domains, adapted to the variable 
environment.  
 
3
 Borrowing the C-terminal domain as a scaffold for engineering a new heme-dependent 
catalyst: 
The work described here also has special bearing on levels two and three of heme protein 
structure and function.  In particular, principles understood from investigations of other heme 
protein structures were applied toward engineering new heme-based catalysts. KatGs have a 
novel two-domain structure, which appears to be the result of a gene duplication and fusion event. 
Taking advantage of this evolutionary history, we borrow the ?peroxidase-like? C-terminal 
domain as a scaffold to engineer a new heme-dependent catalyst, and in particular, restore its 
peroxidative ability. A pathway toward restored peroxidase activity and enzymes with alternative 
catalytic abilities has been laid.  With new proteins competent for association with the cofactor, 
directed evolutionary approaches can now be employed in which the target activity serves as the 
basis for selection of variants from one round of mutagenesis to the next.  
 
4
Diagnosis of the M-Y-W covalent adduct in KatG by SDS-PAGE: 
This dissertation describes a novel technique to monitor an important component of KatG 
structure (M-Y-W covalent adduct). As we demonstrated, formation of this M-Y-W in the KatG 
 172 
adduct required the redox-active porphyrin as the prosthetic group in the presence of oxidant. 
This technique is sensitive to the identity of the cofactor and rightly identifies substitutions and 
deletions that prevent formation of the adduct. Most importantly, for this dissertation, the 
technique demonstrates that part of the active site conformational control exerted by the C-
terminal domain is that which is required to establish the critical M-Y-W adduct. 
 
KatG is a conundrum of protein structure and function.  It has a robust catalase activity that 
is only observed in a class of enzymes to which it has no relation, and its closest relatives which 
share essentially identical active sites have virtually no capacity as catalases. In this regard, KatG 
does possess several novel protein structural components that are to one extent or another 
peripheral to the active site.  This points to a clear model for evaluating how external structures 
exert control over active site conformation and catalytic ability. Among the novel structural 
features of KatG, the C-terminal domain was the first to be recognized. It has the clearest 
evolutionary origin.  It is by far the largest and is situated the greatest distance from the active 
site.  It is also the KatG structure about which the least is understood.  Prior to the initiation of 
the research described in this dissertation, it was known that the C-terminal domain was 
indispensible to the function of KatG, and it showed a very specific impact on the KatG active 
site some 30 ? away. Works described in this dissertation first identify the critical structural 
component (I?-helix) in the C-terminal domain as a platform and discover the mechanism(s) by 
which the C-terminal domain communicates with and controls the active site. These findings 
contribute to answering fundamental questions of enzyme structure and function with clues to 
how new functions emerge in nature through evolutionary processes.  These studies also point 
toward ways that new heme enzymes may be generated and improves our understanding of the 
 173 
structure of a protein underneath antibiotic resistance in Mycobacterium tuberculosis and 
enhanced virulence of pathogenic organisms. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 174 
 
 
 
 
 
REFERENCES 
 
 
[1] L.C. Seaver, J.A. Imlay, Are respiratory enzymes the primary sources of intracellular 
hydrogen peroxide?, J Biol Chem 279 (2004) 48742-48750. 
[2] S. Mishra, J. Imlay, Why do bacteria use so many enzymes to scavenge hydrogen 
peroxide?, Arch Biochem Biophys 525 (2012) 145-160. 
[3] S. Jang, J.A. Imlay, Micromolar intracellular hydrogen peroxide disrupts metabolism by 
damaging iron-sulfur enzymes, J Biol Chem 282 (2007) 929-937. 
[4] J.M. Sobota, J.A. Imlay, Iron enzyme ribulose-5-phosphate 3-epimerase in Escherichia 
coli is rapidly damaged by hydrogen peroxide but can be protected by manganese, P Natl 
Acad Sci USA 108 (2011) 5402-5407. 
[5] M. Zamocky, P.G. Furtmuller, C. Obinger, Evolution of catalases from bacteria to 
humans, Antioxid Redox Signal 10 (2008) 1527-1547. 
[6] M. Zamocky, B. Gasselhuber, P.G. Furtmuller, C. Obinger, Molecular evolution of 
hydrogen peroxide degrading enzymes, Arch Biochem Biophys 525 (2012) 131-144. 
[7] J. Bravo, M.J. Mate, T. Schneider, J. Switala, K. Wilson, P.C. Loewen, I. Fita, Structure 
of catalase HPII from Escherichia coli at 1.9 A resolution, Proteins 34 (1999) 155-166. 
[8] M.S. Sevinc, J. Switala, J. Bravo, I. Fita, P.C. Loewen, Truncation and heme pocket 
mutations reduce production of functional catalase HPII in Escherichia coli, Protein Eng 
11 (1998) 549-555. 
 175 
[9] I. Fita, M.G. Rossmann, The active center of catalase, J Mol Biol 185 (1985) 21-37. 
[10] J. Bravo, N. Verdaguer, J. Tormo, C. Betzel, J. Switala, P.C. Loewen, I. Fita, Crystal 
structure of catalase HPII from Escherichia coli, Structure 3 (1995) 491-502. 
[11] S.P. de Visser, What external perturbations influence the electronic properties of catalase 
compound I?, Inorg Chem 45 (2006) 9551-9557. 
[12] J. Bravo, I. Fita, J.C. Ferrer, W. Ens, A. Hillar, J. Switala, P.C. Loewen, Identification of 
a novel bond between a histidine and the essential tyrosine in catalase HPII of 
Escherichia coli, Protein Sci 6 (1997) 1016-1023. 
[13] S.G. Kalko, J.L. Gelpi, I. Fita, M. Orozco, Theoretical study of the mechanisms of 
substrate recognition by catalase, J Am Chem Soc 123 (2001) 9665-9672. 
[14] P, Nichols, I, Fita, P.C. Lowen. Enzymology and structure of catalases. In Advances in 
Inorganic Chemistry, Sykes, Academic Press, New York, 2001 
[15] M. Alfonso-Prieto, X. Biarnes, P. Vidossich, C. Rovira, The molecular mechanism of the 
catalase reaction, J Am Chem Soc 131 (2009) 11751-11761. 
[16] S. Kato, T. Ueno, S. Fukuzumi, Y. Watanabe, Catalase reaction by myoglobin mutants 
and native catalase: mechanistic investigation by kinetic isotope effect, J Biol Chem 279 
(2004) 52376-52381. 
[17] J.T. Groves, G.A. McClusky, Aliphatic hydroxylation by highly purified liver 
microsomal cytochrome P-450. Evidence for a carbon radical intermediate, Biochem 
Biophys Res Commun 81 (1978) 154-160. 
[18] M. Zamocky, C. Jakopitsch, P.G. Furtmuller, C. Dunand, C. Obinger, The peroxidase-
cyclooxygenase superfamily: Reconstructed evolution of critical enzymes of the innate 
immune system, Proteins 72 (2008) 589-605. 
 176 
[19] F. Passardi, N. Bakalovic, F.K. Teixeira, M. Margis-Pinheiro, C. Penel, C. Dunand, 
Prokaryotic origins of the non-animal peroxidase superfamily and organelle-mediated 
transmission to eukaryotes, Genomics 89 (2007) 567-579. 
[20] D. Koua, L. Cerutti, L. Falquet, C.J.A. Sigrist, G. Theiler, N. Hulo, C. Dunand, 
PeroxiBase: a database with new tools for peroxidase family classification, Nucleic Acids 
Res 37 (2009) D261-D266. 
[21] E. Torres, M. Ayala, Biocatalysis based on heme peroxidases, Chapter 2, Molecular 
Phylogeny of heme peroxidases, Page 7-36,  2010,  Springer,  New York 
[22] N. Ellfolk, R. Soininen, Pseudomonas cytochrome c peroxidase. I. Purification 
procedure, Acta Chem Scand 24 (1970) 2126-2136. 
[23] A. Echalier, T. Brittain, J. Wright, S. Boycheva, G.B. Mortuza, V. Fulop, N.J. 
Watmough, Redox-linked structural changes associated with the formation of a 
catalytically competent form of the diheme cytochrome c peroxidase from Pseudomonas 
aeruginosa, Biochemistry-Us 47 (2008) 1947-1956. 
[24] A. Echalier, C.F. Goodhew, G.W. Pettigrew, V. Fulop, Activation and catalysis of the di-
heme cytochrome c peroxidase from Paracoccus pantotrophus, Structure 14 (2006) 107-
117. 
[25] N. Ellfolk, M. Ronnberg, R. Aasa, L.E. Andreasson, T. Vanngard, Properties and 
function of the two hemes in Pseudomonas cytochrome c peroxidase, Biochim Biophys 
Acta 743 (1983) 23-30. 
[26] R. Gilmour, C.F. Goodhew, G.W. Pettigrew, S. Prazeres, I. Moura, J.J. Moura, 
Spectroscopic characterization of cytochrome c peroxidase from Paracoccus 
denitrificans, Biochem J 294 ( Pt 3) (1993) 745-752. 
 177 
[27] V. Fulop, C.J. Ridout, C. Greenwood, J. Hajdu, Crystal structure of the di-haem 
cytochrome c peroxidase from Pseudomonas aeruginosa, Structure 3 (1995) 1225-1233. 
[28] L. De Smet, S.N. Savvides, E. Van Horen, G. Pettigrew, J.J. Van Beeumen, Structural 
and mutagenesis studies on the cytochrome c peroxidase from Rhodobacter capsulatus 
provide new insights into structure-function relationships of bacterial di-heme 
peroxidases, J Biol Chem 281 (2006) 4371-4379. 
[29] M. Hoffmann, J. Seidel, O. Einsle, CcpA from Geobacter sulfurreducens Is a Basic Di-
Heme Cytochrome c Peroxidase, J Mol Biol 393 (2009) 951-965. 
[30] H. Shimizu, D.J. Schuller, W.N. Lanzilotta, M. Sundaramoorthy, D.M. Arciero, A.B. 
Hooper, T.L. Poulos, Crystal structure of Nitrosomonas europaea cytochrome c 
peroxidase and the structural basis for ligand switching in bacterial di-heme peroxidases, 
Biochemistry-Us 40 (2001) 13483-13490. 
[31] J.M. Dias, T. Alves, C. Bonifacio, A.S. Pereira, J. Trincao, D. Bourgeois, I. Moura, M.J. 
Romao, Structural basis for the mechanism of Ca(2+) activation of the di-heme 
cytochrome c peroxidase from Pseudomonas nautica 617, Structure 12 (2004) 961-973. 
[32] G.W. Pettigrew, A. Echalier, S.R. Pauleta, Structure and mechanism in the bacterial 
dihaem cytochrome c peroxidases, J Inorg Biochem 100 (2006) 551-567. 
[33] R. Gilmour, C.F. Goodhew, G.W. Pettigrew, S. Prazeres, J.J. Moura, I. Moura, The 
kinetics of the oxidation of cytochrome c by Paracoccus cytochrome c peroxidase, 
Biochem J 300 ( Pt 3) (1994) 907-914. 
[34] T. Sato, S. Hara, T. Matsui, G. Sazaki, S. Saijo, T. Ganbe, N. Tanaka, Y. Sugano, M. 
Shoda, A unique dye-decolorizing peroxidase, DyP, from Thanatephorus cucumeris Dec 
 178 
1: heterologous expression, crystallization and preliminary X-ray analysis, Acta 
Crystallogr D Biol Crystallogr 60 (2004) 149-152. 
[35] Y. Sugano, R. Muramatsu, A. Ichiyanagi, T. Sato, M. Shoda, DyP, a unique dye-
decolorizing peroxidase, represents a novel heme peroxidase family: ASP171 replaces 
the distal histidine of classical peroxidases, J Biol Chem 282 (2007) 36652-36658. 
[36] H.J. Ogola, N. Hashimoto, S. Miyabe, H. Ashida, T. Ishikawa, H. Shibata, Y. Sawa, 
Enhancement of hydrogen peroxide stability of a novel Anabaena sp. DyP-type 
peroxidase by site-directed mutagenesis of methionine residues, Appl Microbiol 
Biotechnol 87 (2010) 1727-1736. 
[37] Y. Sugano, Y. Matsushima, K. Tsuchiya, H. Aoki, M. Hirai, M. Shoda, Degradation 
pathway of an anthraquinone dye catalyzed by a unique peroxidase DyP from 
Thanatephorus cucumeris Dec 1, Biodegradation 20 (2009) 433-440. 
[38] K.G. Welinder, J.M. Mauro, L. Norskov-Lauritsen, Structure of plant and fungal 
peroxidases, Biochem Soc Trans 20 (1992) 337-340. 
[39] T. Yoshida, H. Tsuge, H. Konno, T. Hisabori, Y. Sugano, The catalytic mechanism of 
dye-decolorizing peroxidase DyP may require the swinging movement of an aspartic acid 
residue, The FEBS journal 278 (2011) 2387-2394. 
[40] C. Zubieta, R. Joseph, S.S. Krishna, D. McMullan, M. Kapoor, H.L. Axelrod, M.D. 
Miller, P. Abdubek, C. Acosta, T. Astakhova, D. Carlton, H.J. Chiu, T. Clayton, M.C. 
Deller, L. Duan, Y. Elias, M.A. Elsliger, J. Feuerhelm, S.K. Grzechnik, J. Hale, G.W. 
Han, L. Jaroszewski, K.K. Jin, H.E. Klock, M.W. Knuth, P. Kozbial, A. Kumar, D. 
Marciano, A.T. Morse, K.D. Murphy, E. Nigoghossian, L. Okach, S. Oommachen, R. 
Reyes, C.L. Rife, P. Schimmel, C.V. Trout, H. van den Bedem, D. Weekes, A. White, Q. 
 179 
Xu, K.O. Hodgson, J. Wooley, A.M. Deacon, A. Godzik, S.A. Lesley, I.A. Wilson, 
Identification and structural characterization of heme binding in a novel dye-decolorizing 
peroxidase, TyrA, Proteins 69 (2007) 234-243. 
[41] M. Hofrichter, R. Ullrich, Heme-thiolate haloperoxidases: versatile biocatalysts with 
biotechnological and environmental significance, Appl Microbiol Biotechnol 71 (2006) 
276-288. 
[42] J.R. Beckwith, L.P. Hager, Biological Chlorination. Viii. Late Intermediates in the 
biosynthesis of caldariomycin, J Biol Chem 238 (1963) 3091-3094. 
[43] J.R. Beckwith, R. Clark, L.P. Hager, Biological Chlorination. Vii. Studies on the 
biosynthesis of caldariomycin, J Biol Chem 238 (1963) 3086-3090. 
[44] L.P. Hager, D.R. Morris, F.S. Brown, H. Eberwein, Chloroperoxidase. II. Utilization of 
halogen anions, J Biol Chem 241 (1966) 1769-1777. 
[45] E.L. Patterson, W.W. Andres, L.A. Mitscher, Isolation of the bromo analogue of 
caldariomycin from Caldariomyces fumago, Appl Microbiol 15 (1967) 528-530. 
[46] N. Itoh, Y. Izumi, H. Yamada, Characterization of nonheme iron and reaction mechanism 
of bromoperoxidase in Corallina pilulifera, J Biol Chem 262 (1987) 11982-11987. 
[47] K. Kuhnel, W. Blankenfeldt, J. Terner, I. Schlichting, Crystal structures of 
chloroperoxidase with its bound substrates and complexed with formate, acetate, and 
nitrate, J Biol Chem 281 (2006) 23990-23998. 
[48] M. Sundaramoorthy, J. Terner, T.L. Poulos, The crystal structure of chloroperoxidase: a 
heme peroxidase--cytochrome P450 functional hybrid, Structure 3 (1995) 1367-1377. 
[49] E. Malle, P.G. Furtmuller, W. Sattler, C. Obinger, Myeloperoxidase: a target for new 
drug development?, Br J Pharmacol 152 (2007) 838-854. 
 180 
[50] D.T. Akin, J.M. Kinkade, Jr., Processing of a newly identified intermediate of human 
myeloperoxidase in isolated granules occurs at neutral pH, J Biol Chem 261 (1986) 8370-
8375. 
[51] H.P. Koeffler, J. Ranyard, M. Pertcheck, Myeloperoxidase: its structure and expression 
during myeloid differentiation, Blood 65 (1985) 484-491. 
[52] R. Wever, H. Plat, Spectral properties of myeloperoxidase and its ligand complexes, 
Biochim Biophys Acta 661 (1981) 235-239. 
[53] J. Arnhold, E. Monzani, P.G. Furtmuller, M. Zederbauer, L. Casella, C. Obinger, Kinetics 
and thermodynamics of halide and nitrite oxidation by mammalian heme peroxidases, 
Eur J Inorg Chem  (2006) 3801-3811. 
[54] T.J. Fiedler, C.A. Davey, R.E. Fenna, X-ray crystal structure and characterization of 
halide-binding sites of human myeloperoxidase at 1.8 A resolution, J Biol Chem 275 
(2000) 11964-11971. 
[55] S. Shigeoka, T. Ishikawa, M. Tamoi, Y. Miyagawa, T. Takeda, Y. Yabuta, K. 
Yoshimura, Regulation and function of ascorbate peroxidase isoenzymes, J Exp Bot 53 
(2002) 1305-1319. 
[56] V.P. Skulachev, Cytochrome c in the apoptotic and antioxidant cascades, FEBS Lett 423 
(1998) 275-280. 
[57] J.E. Erman, L.B. Vitello, Yeast cytochrome c peroxidase: mechanistic studies via protein 
engineering, Biochim Biophys Acta 1597 (2002) 193-220. 
[58] A.T. Martinez, M. Speranza, F.J. Ruiz-Duenas, P. Ferreira, S. Camarero, F. Guillen, M.J. 
Martinez, A. Gutierrez, J.C. del Rio, Biodegradation of lignocellulosics: microbial, 
 181 
chemical, and enzymatic aspects of the fungal attack of lignin, Int Microbiol 8 (2005) 
195-204. 
[59] K. Piontek, A.T. Smith, W. Blodig, Lignin peroxidase structure and function, Biochem 
Soc Trans 29 (2001) 111-116. 
[60] F. Passardi, D. Longet, C. Penel, C. Dunand, The class III peroxidase multigenic family 
in rice and its evolution in land plants, Phytochemistry 65 (2004) 1879-1893. 
[61] M.A. Bernards, W.D. Fleming, D.B. Llewellyn, R. Priefer, X. Yang, A. Sabatino, G.L. 
Plourde, Biochemical characterization of the suberization-associated anionic peroxidase 
of potato, Plant Physiol 121 (1999) 135-146. 
[62] S.D. Allison, J.C. Schultz, Differential activity of peroxidase isozymes in response to 
wounding, gypsy moth, and plant hormones in northern red oak (Quercus rubra L.), J 
Chem Ecol 30 (2004) 1363-1379. 
[63] I.G. Gazaryan, L.M. Lagrimini, G.A. Ashby, R.N. Thorneley, Mechanism of indole-3-
acetic acid oxidation by plant peroxidases: anaerobic stopped-flow spectrophotometric 
studies on horseradish and tobacco peroxidases, Biochem J 313 ( Pt 3) (1996) 841-847. 
[64] C. Mathe, A. Barre, C. Jourda, C. Dunand, Evolution and expression of class III 
peroxidases, Arch Biochem Biophys 500 (2010) 58-65. 
[65] M. Gajhede, D.J. Schuller, A. Henriksen, A.T. Smith, T.L. Poulos, Crystal structure of 
horseradish peroxidase C at 2.15 A resolution, Nat Struct Biol 4 (1997) 1032-1038. 
[66] T.G. Douroupi, I.S. Papassideri, D.J. Stravopodis, L.H. Margaritis, Molecular cloning 
and tissue-specific transcriptional regulation of the first peroxidase family member, Gene 
362 (2005) 57-69. 
 182 
[67] B. Lige, S. Ma, R.B. van Huystee, The effects of the site-directed removal of N-
glycosylation from cationic peanut peroxidase on its function, Arch Biochem Biophys 
386 (2001) 17-24. 
[68] C. Gabaldon, L.V. Gomez-Ros, M.J. Nunez-Flores, A. Esteban-Carrasco, A.R. Barcelo, 
Post-translational modifications of the basic peroxidase isoenzyme from Zinnia elegans, 
Plant Mol Biol 65 (2007) 43-61. 
[69] N.C. Veitch, Horseradish peroxidase: a modern view of a classic enzyme, 
Phytochemistry 65 (2004) 249-259. 
[70] S. Nagano, M. Tanaka, K. Ishimori, Y. Watanabe, I. Morishima, Catalytic roles of the 
distal site asparagine-histidine couple in peroxidases, Biochemistry-Us 35 (1996) 14251-
14258. 
[71] H.S. Pappa, A.E. Cass, A step towards understanding the folding mechanism of 
horseradish peroxidase. Tryptophan fluorescence and circular dichroism equilibrium 
studies, Eur J Biochem 6 / FEBS 212 (1993) 227-235. 
[72] A.T. Smith, N. Santama, S. Dacey, M. Edwards, R.C. Bray, R.N. Thorneley, J.F. Burke, 
Expression of a synthetic gene for horseradish peroxidase C in Escherichia coli and 
folding and activation of the recombinant enzyme with Ca
2+
 and heme, J Biol Chem 265 
(1990) 13335-13343. 
[73] H.B. Dunford., 1999. Heme peroxidases. Wiley-VCH, New York 
[74] N.C.Veitch., Smith, A.T. 2001. Horseradish peroxidase. Adv. Inorg. Chem. 51, 107-162 
[75] G.I. Berglund, G.H. Carlsson, A.T. Smith, H. Szoke, A. Henriksen, J. Hajdu, The 
catalytic pathway of horseradish peroxidase at high resolution, Nature 417 (2002) 463-
468. 
 183 
[76] T.K. Kirk, R.L. Farrell, Enzymatic "combustion": the microbial degradation of lignin, 
Annu Rev Microbiol 41 (1987) 465-505. 
[77] D.W. Wong, Structure and action mechanism of ligninolytic enzymes, Appl Biochem 
Biotechnol 157 (2009) 174-209. 
[78] K.E. Hammel, K.A. Jensen, Jr., M.D. Mozuch, L.L. Landucci, M. Tien, E.A. Pease, 
Ligninolysis by a purified lignin peroxidase, J Biol Chem 268 (1993) 12274-12281. 
[79] M. Tien, T.K. Kirk, Lignin-Degrading Enzyme from the Hymenomycete Phanerochaete 
chrysosporium Burds, Science 221 (1983) 661-663. 
[80] K. Valli, H. Wariishi, M.H. Gold, Oxidation of monomethoxylated aromatic compounds 
by lignin peroxidase: role of veratryl alcohol in lignin biodegradation, Biochemistry-Us 
29 (1990) 8535-8539. 
[81] T.L. Poulos, S.L. Edwards, H. Wariishi, M.H. Gold, Crystallographic refinement of 
lignin peroxidase at 2 A, J Biol Chem 268 (1993) 4429-4440. 
[82] W. Blodig, W.A. Doyle, A.T. Smith, K. Winterhalter, T. Choinowski, K. Piontek, 
Autocatalytic formation of a hydroxy group at C? of trp171 in lignin peroxidase, 
Biochemistry-Us 37 (1998) 8832-8838. 
[83] W.A. Doyle, W. Blodig, N.C. Veitch, K. Piontek, A.T. Smith, Two substrate interaction 
sites in lignin peroxidase revealed by site-directed mutagenesis, Biochemistry-Us 37 
(1998) 15097-15105. 
[84] W. Blodig, A.T. Smith, W.A. Doyle, K. Piontek, Crystal structures of pristine and 
oxidatively processed lignin peroxidase expressed in Escherichia coli and of the W171F 
variant that eliminates the redox active tryptophan 171. Implications for the reaction 
mechanism, J Mol Biol 305 (2001) 851-861. 
 184 
[85] C.D. Millis, D.Y. Cai, M.T. Stankovich, M. Tien, Oxidation-reduction potentials and 
ionization states of extracellular peroxidases from the lignin-degrading fungus 
Phanerochaete chrysosporium, Biochemistry-Us 28 (1989) 8484-8489. 
[86] M. Sundaramoorthy, K. Kishi, M.H. Gold, T.L. Poulos, Preliminary crystallographic 
analysis of manganese peroxidase from Phanerochaete chrysosporium, J Mol Biol 238 
(1994) 845-848. 
[87] M. Sundaramoorthy, K. Kishi, M.H. Gold, T.L. Poulos, Crystal structures of substrate 
binding site mutants of manganese peroxidase, J Biol Chem 272 (1997) 17574-17580. 
[88] M. Perez-Boada, F.J. Ruiz-Duenas, R. Pogni, R. Basosi, T. Choinowski, M.J. Martinez, 
K. Piontek, A.T. Martinez, Versatile peroxidase oxidation of high redox potential 
aromatic compounds: site-directed mutagenesis, spectroscopic and crystallographic 
investigation of three long-range electron transfer pathways, J Mol Biol 354 (2005) 385-
402. 
[89] R. Pogni, M.C. Baratto, C. Teutloff, S. Giansanti, F.J. Ruiz-Duenas, T. Choinowski, K. 
Piontek, A.T. Martinez, F. Lendzian, R. Basosi, A tryptophan neutral radical in the 
oxidized state of versatile peroxidase from Pleurotus eryngii: a combined multifrequency 
EPR and density functional theory study, J Biol Chem 281 (2006) 9517-9526. 
[90] M.H. Gold, H. Wariishi, K. Valli, Extracellular Peroxidases Involved in Lignin 
Degradation by the White Rot Basidiomycete Phanerochaete-Chrysosporium, Acs Sym 
Ser 389 (1989) 127-140. 
[91] G.V. Reddy, M. Sridhar, M.H. Gold, Cleavage of nonphenolic beta-1 diarylpropane 
lignin model dimers by manganese peroxidase from Phanerochaete chrysosporium, Eur J 
Biochem / FEBS 270 (2003) 284-292. 
 185 
[92] M. Zamocky, P.G. Furtmuller, C. Obinger, Evolution of structure and function of Class I 
peroxidases, Arch Biochem Biophys 500 (2010) 45-57. 
[93] M.G. Klotz, P.C. Loewen, The molecular evolution of catalatic hydroperoxidases: 
evidence for multiple lateral transfer of genes between prokaryota and from bacteria into 
eukaryota, Mol Biol Evol 20 (2003) 1098-1112. 
[94] I. Morgenstern, S. Klopman, D.S. Hibbett, Molecular evolution and diversity of lignin 
degrading heme peroxidases in the Agaricomycetes, J Mol Evol 66 (2008) 243-257. 
[95] L.A. del Rio, F.J. Corpas, L.M. Sandalio, J.M. Palma, M. Gomez, J.B. Barroso, Reactive 
oxygen species, antioxidant systems and nitric oxide in peroxisomes, J Exp Bot 53 (2002) 
1255-1272. 
[96] A. Boveris, Acta Physiol. Latin. Am. 26 (1976) 31-37 
[97] D.V. Dabir, E.P. Leverich, S.K. Kim, F.D. Tsai, M. Hirasawa, D.B. Knaff, C.M. Koehler, 
A role for cytochrome c and cytochrome c peroxidase in electron shuttling from Erv1, 
Embo J 26 (2007) 4801-4811. 
[98] S.S. Giles, J.R. Perfect, G.M. Cox, Cytochrome c peroxidase contributes to the 
antioxidant defense of Cryptococcus neoformans, Fungal Genet Biol 42 (2005) 20-29. 
[99] C.A. Bonagura, B. Bhaskar, H. Shimizu, H. Li, M. Sundaramoorthy, D.E. McRee, D.B. 
Goodin, T.L. Poulos, High-resolution crystal structures and spectroscopy of native and 
compound I cytochrome c peroxidase, Biochemistry-Us 42 (2003) 5600-5608. 
[100] T.L. Poulos, Thirty years of heme peroxidase structural biology, Arch Biochem Biophys 
500 (2010) 3-12. 
[101] T.P. Barrows, B. Bhaskar, T.L. Poulos, Electrostatic control of the tryptophan radical in 
cytochrome c peroxidase, Biochemistry-Us 43 (2004) 8826-8834. 
 186 
[102] M. Zamocky, Phylogenetic relationships in class I of the superfamily of bacterial, fungal, 
and plant peroxidases, Eur J Biochem 271 (2004) 3297-3309. 
[103] B.C. Finzel, T.L. Poulos, J. Kraut, Crystal structure of yeast cytochrome c peroxidase 
refined at 1.7-A resolution, J Biol Chem 259 (1984) 13027-13036. 
[104] K.H. Sharp, M. Mewies, P.C. Moody, E.L. Raven, Crystal structure of the ascorbate 
peroxidase-ascorbate complex, Nat Struct Biol 10 (2003) 303-307. 
[105] Y. Zhang, B. Heym, B. Allen, D. Young, S. Cole, The catalase-peroxidase gene and 
isoniazid resistance of Mycobacterium tuberculosis, Nature 358 (1992) 591-593. 
[106] S. Chouchane, I. Lippai, R.S. Magliozzo, Catalase-peroxidase (Mycobacterium 
tuberculosis KatG) catalysis and isoniazid activation, Biochemistry-Us 39 (2000) 9975-
9983. 
[107] B.F. Lei, C.J. Wei, S.C. Tu, Action mechanism of antitubercular isoniazid - Activation 
mycobacterium tuberculosis KatG, isolation, and characterization of InhA inhibitor, J 
Biol Chem 275 (2000) 2520-2526. 
[108] C. Vilcheze, W.R. Jacobs, Jr., The mechanism of isoniazid killing: clarity through the 
scope of genetics, Annu Rev Microbiol 61 (2007) 35-50. 
[109] X.B. Zhao, H. Yu, S.W. Yu, F. Wang, J.C. Sacchettini, R.S. Magliozzo, Hydrogen 
peroxide-mediated isoniazid activation catalyzed by Mycobacterium tuberculosis 
catalase-peroxidase (KatG) and its S315T mutant, Biochemistry-Us 45 (2006) 4131-
4140. 
[110] D.E. Minnikin, L. Kremer, L.G. Dover, G.S. Besra, The methyl-branched fortifications of 
Mycobacterium tuberculosis, Chem Biol 9 (2002) 545-553. 
 187 
[111] E. Garcia, Y.A. Nedialkov, J. Elliott, V.L. Motin, R.R. Brubaker, Molecular 
characterization of KatY (Antigen 5), a thermoregulated chromosomally encoded 
catalase-peroxidase of Yersinia pestis, J Bacteriol 181 (1999) 3114-3122. 
[112] R.J. Mehigh, R.R. Braubaker, Major stable peptides of Yersinia pestis synthesized during 
the low-calcium response, Infect Immun 61 (1993) 13-22. 
[113] B.A. Chromy, M.W. Choi, G.A. Murphy, A.D. Gonzales, C.H. Corzett, B.C. Chang, J.P. 
Fitch, S.L. McCutchen-Maloney, Proteomic characterization of Yersinia pestis virulence, 
J Bacteriol 187 (2005) 8172-8180. 
[114] P. Bandyopadhyay, B. Byrne, Y. Chan, M.S. Swanson, H.M. Steinman, Legionella 
pneumophila catalase-peroxidases are required for proper trafficking and growth in 
primary macrophages, Infect Immun 71 (2003) 4526-4535. 
[115] G.A. Uhlich, C.Y. Chen, B.J. Cottrell, P.L. Irwin, J.G. Phillips, Peroxide resistance in 
Escherichia coli serotype O157 : H7 biofilms is regulated by both RpoS-dependent and -
independent mechanisms, Microbiology 158 (2012) 2225-2234. 
[116] G.A. Uhlich, KatP contributes to OxyR-regulated hydrogen peroxide resistance in 
Escherichia coli serotype O157 : H7, Microbiology 155 (2009) 3589-3598. 
[117] V. Cannac-Caffrey, G. Hudry-Clergeon, Y. Petillot, J. Gagnon, G. Zaccai, B. Franzetti, 
The protein sequence of an archaeal catalase-peroxidase, Biochimie 80 (1998) 1003-
1011. 
[118] K.G. Welinder, Bacterial catalase-peroxidases are gene duplicated members of the plant 
peroxidase superfamily, Biochim Biophys Acta 1080 (1991) 215-220. 
[119] Y. Yamada, T. Fujiwara, T. Sato, N. Igarashi, N. Tanaka, The 2.0 ? crystal structure of 
catalase-peroxidase from Haloarcula marismortui, Nat Struct Biol 9 (2002) 691-695. 
 188 
[120] K. Wada, T. Tada, Y. Nakamura, T. Kinoshita, M. Tamoi, S. Shigeoka, K. Nishimura, 
Crystallization and preliminary X-ray diffraction studies of catalase-peroxidase from 
Synechococcus PCC 7942, Acta Crystallogr D Biol Crystallogr 58 (2002) 157-159. 
[121] X. Carpena, S. Loprasert, S. Mongkolsuk, J. Switala, P.C. Loewen, I. Fita, Catalase-
peroxidase KatG of Burkholderia pseudomallei at 1.7 ? resolution, J Mol Biol 327 
(2003) 475-489. 
[122] T. Bertrand, N.A. Eady, J.N. Jones, Jesmin, J.M. Nagy, B. Jamart-Gregoire, E.L. Raven, 
K.A. Brown, Crystal structure of Mycobacterium tuberculosis catalase-peroxidase, J Biol 
Chem 279 (2004) 38991-38999. 
[123] H.A. Heering, C. Indiani, G. Regelsberger, C. Jakopitsch, C. Obinger, G. Smulevich, 
New insights into the heme cavity structure of catalase-peroxidase: A spectroscopic 
approach to the recombinant Synechocystis enzyme and selected distal cavity mutants, 
Biochemistry-Us 41 (2002) 9237-9247. 
[124] C. Jakopitsch, M. Auer, G. Regelsberger, W. Jantschko, P.G. Furtmuller, F. Ruker, C. 
Obinger, The catalytic role of the distal site asparagine-histidine couple in catalase-
peroxidases, Eur J Biochem 270 (2003) 1006-1013. 
[125] C. Jakopitsch, M. Auer, G. Regelsberger, W. Jantschko, P.G. Furtmuller, F. Ruker, C. 
Obinger, Distal site aspartate is essential in the catalase activity of catalase-peroxidases, 
Biochemistry-Us 42 (2003) 5292-5300. 
[126] C. Jakopitsch, E. Droghetti, F. Schmuckenschlager, P.G. Furtmuller, G. Smulevich, C. 
Obinger, Role of the main access channel of catalase-peroxidase in catalysis, J Biol 
Chem 280 (2005) 42411-42422. 
 189 
[127] M. Zamocky, Q. Garcia-Fernandez, B. Gasselhuber, C. Jakopitsch, P.G. Furtmuller, P.C. 
Loewen, I. Fita, C. Obinger, X. Carpena, High conformational stability of secreted 
eukaryotic catalase-peroxidases -answers from first crystal structure and unfolding 
studies, J Biol Chem  2012. 
[128] M. Zamocky, G. Regelsberger, C. Jakopitsch, C. Obinger, The molecular peculiarities of 
catalase-peroxidases, FEBS Lett 492 (2001) 177-182. 
[129] Y.J. Li, D.C. Goodwin, Vital roles of an interhelical insertion in catalase-peroxidase 
bifunctionality, Biochem Biophys Res Commun 318 (2004) 970-976. 
[130] S. N. Kudalkar, Roles of large loops in catalytic versatility of catalase-peroxidases: 
significance of peripheral structures in improvising enzyme functions (2012) 
Dissertation, Auburn University. 
[131] T. Bertrand, N.A.J. Eady, J.N. Jones, J.M. Nagy, B. Jamart-Gregoire, E.L. Raven, K.A. 
Brown, Crystal structure of Mycobacterium tuberculosis catalase-peroxidase, J Biol 
Chem 279 (2004) 38991-38999. 
[132] R.A. Ghiladi, G.M. Knudsen, K.F. Medzihradszky, P.R. Ortiz de Montellano, The Met-
Tyr-Trp cross-link in Mycobacterium tuberculosis catalase-peroxidase (KatG): 
autocatalytic formation and effect on enzyme catalysis and spectroscopic properties, J 
Biol Chem 280 (2005) 22651-22663. 
[133] C. Jakopitsch, D. Kolarich, G. Petutschnig, P.G. Furtmuller, C. Obinger, Distal side 
tryptophan, tyrosine and methionine in catalase-peroxidases are covalently linked in 
solution, FEBS Lett 552 (2003) 135-140. 
[134] L.J. Donald, O.V. Krokhin, H.W. Duckworth, B. Wiseman, T. Deemagarn, R. Singh, J. 
Switala, X. Carpena, I. Fita, P.C. Loewen, Characterization of the catalase-peroxidase 
 190 
KatG from Burkholderia pseudomallei by mass spectrometry, J Biol Chem 278 (2003) 
35687-35692. 
[135] R.A. Ghiladi, K.F. Medzihradszky, P.R. Ortiz de Montellano, Role of the Met-Tyr-Trp 
cross-link in Mycobacterium tuberculosis catalase-peroxidase (KatG) as revealed by 
KatG(M255I), Biochemistry-Us 44 (2005) 15093-15105. 
[136] S.N. Kudalkar, R.A. Campbell, Y. Li, C.L. Varnado, C. Prescott, D.C. Goodwin, 
Enhancing the peroxidatic activity of KatG by deletion mutagenesis, J Inorg Biochem 
116C (2012) 106-115. 
[137] J. Vlasits, C. Jakopitsch, M. Bernroitner, M. Zamocky, P.G. Furtmuller, C. Obinger, 
Mechanisms of catalase activity of heme peroxidases, Arch Biochem Biophys 500 (2010) 
74-81. 
[138] X.B. Zhao, J. Suarez, A. Khajo, S.W. Yu, L. Metlitsky, R.S. Magliozzo, A radical on the 
met-tyr-trp modification required for catalase activity in catalase-peroxidase is 
established by isotopic labeling and site-directed mutagenesis, J Am Chem Soc 132 
(2010) 8268. 
[139] J. Suarez, K. Ranguelova, A.A. Jarzecki, J. Manzerova, V. Krymov, X.B. Zhao, S.W. Yu, 
L. Metlitsky, G.J. Gerfen, R.S. Magliozzo, An oxyferrous heme/protein-based radical 
intermediate is catalytically competent in the catalase reaction of mycobacterium 
tuberculosis catalase-peroxidase (KatG), J Biol Chem 284 (2009) 7017-7029. 
[140] C. Jakopitsch, J. Vlasits, B. Wiseman, P.C. Loewen, C. Obinger, Redox intermediates in 
the catalase cycle of catalase-peroxidases from Synechocystis PCC 6803, Burkholderia 
pseudomallei, and Mycobacterium tuberculosis, Biochemistry-Us 46 (2007) 1183-1193. 
 191 
[141] M. Zamocky, Phylogenetic relationships in class I of the superfamily of bacterial, fungal, 
and plant peroxidases, Eur J Biochem / FEBS 271 (2004) 3297-3309. 
[142] R.D. Baker, C.O. Cook, D.C. Goodwin, Properties of catalase-peroxidase lacking its C-
terminal domain, Biochem Biophys Res Commun 320 (2004) 833-839. 
[143] X. Carpena, W. Melik-Adamyan, P.C. Loewen, I. Fita, Structure of the C-terminal 
domain of the catalase-peroxidase KatG from Escherichia coli, Acta Crystallogr D Biol 
Crystallogr 60 (2004) 1824-1832. 
[144] R.D. Baker, C.O. Cook, D.C. Goodwin, Catalase-peroxidase active site restructuring by a 
distant and "inactive" domain, Biochemistry-Us 45 (2006) 7113-7121. 
[145] S. Banerjee, M. Zamocky, P.G. Furtmuller, C. Obinger, Probing the two-domain structure 
of homodimeric prokaryotic and eukaryotic catalase-peroxidases, Biochim Biophys Acta 
1804 (2010) 2136-2145. 
[146] X. Carpena, J. Switala, S. Loprasert, S. Mongkolsuk, I. Fita, P.C. Loewen, Crystallization 
and preliminary X-ray analysis of the catalase-peroxidase KatG from Burkholderia 
pseudomallei, Acta Crystallogr D Biol Crystallogr 58 (2002) 2184-2186. 
[147] J.A. Imlay, Cellular defenses against superoxide and hydrogen peroxide, Annu Rev 
Biochem 77 (2008) 755-776. 
[148] P.C. Loewen, Bacterial catalases, Oxidative stress and the molecular biology of 
antioxidant defenses, Cold Spring Harbor Laboratory Press, New York, 1997 
[149] C. Jakopitsch, M. Auer, A. Ivancich, F. Ruker, P.G. Furtmuller, C. Obinger, Total 
conversion of bifunctional catalase-peroxidase (KatG) to monofunctional peroxidase by 
exchange of a conserved distal side tyrosine, J Biol Chem 278 (2003) 20185-20191. 
 192 
[150] S. Yu, S. Girotto, X. Zhao, R.S. Magliozzo, Rapid formation of compound II and a 
tyrosyl radical in the Y229F mutant of Mycobacterium tuberculosis catalase-peroxidase 
disrupts catalase but not peroxidase function, J Biol Chem 278 (2003) 44121-44127. 
[151] Y. Li, D.C. Goodwin, Vital roles of an interhelical insertion in catalase-peroxidase 
bifunctionality, Biochem Biophys Res Commun 318 (2004) 970-976. 
[152] S. More. "round-the-horn site-directed mutagensis-Openwetware, 
http://openwetware.org/wiki/%27Round-the-horn_site-directed_mutagenesis 
[153] C.L. Varnado, K.M. Hertwig, R. Thomas, J.K. Roberts, D.C. Goodwin, Properties of a 
novel periplasmic catalase-peroxidase from Escherichia coli O157 : H7, Arch Biochem 
Biophys 421 (2004) 166-174. 
[154] S.C. Gill, P.H. von Hippel, Calculation of protein extinction coefficients from amino acid 
sequence data, Anal Biochem 182 (1989) 319-326. 
[155] J.E. Falk, Chromatography of Porphyrins and Metalloporphyrins, J Chromatogr 5 (1961) 
277. 
[156] S.L. Scott, W.J. Chen, A. Bakac, J.H. Espenson, Spectroscopic parameters, electrode-
potentials, acid ionization-constants, and electron-exchange rates of the 2,2'-azinobis(3-
ethylbenzothiazoline-6-sulfonate) radicals and ions, J Phys Chem-Us 97 (1993) 6710-
6714. 
[157] D.P. Nelson, L.A. Kiesow, Enthalpy of decomposition of hydrogen-peroxide by catalase 
at 25 degrees c, Anal Biochem 49 (1972) 474. 
[158] A. Fersht, Structure and Mechanism in Protein Science: a Guide to Enzyme Catalysis and 
Proyein Folding. W.H. Freeman and Company, New York, 1999 
[159] W.R.Hagen, Biomolecular EPR Spectroscopy, CRC Press, Boca Raton, FL, 2009 
 193 
[160] W.R. Hagen,TU Delft: Biomolecular EPR Spectroscopy, 
            http://www.bt.tudelft.nl/biomolecularEPRspectroscopy (2012) 
[161] S.E. Jackson, A.R. Fersht, Folding of Chymotrypsin Inhibitor-2 .1. Evidence for a 2-State 
Transition, Biochemistry-Us 30 (1991) 10428-10435. 
[162] C. Zentz, S. Pin, B. Alpert, Stationary and time-resolved circular-dichroism of 
hemoglobins, Method Enzymol 232 (1994) 247-266. 
[163] G. Tsaprailis, D.W.S. Chan, A.M. English, Conformational states in denaturants of 
cytochrome c and horseradish peroxidases examined by fluorescence and circular 
dichroism, Biochemistry-Us 37 (1998) 2004-2016. 
[164] M.R. Cheesman, C. Greenwood, A.J. Thomson, Magnetic Circular-Dichroism of 
Hemoproteins, Adv Inorg Chem Rad 36 (1991) 201-255. 
[165] A. Ivancich, C. Jakopitsch, M. Auer, S. Un, C. Obinger, Protein-based radicals in the 
catalase-peroxidase of Synechocystis PCC6803: A multifrequency EPR investigation of 
wild-type and variants on the environment of the heme active site, J Am Chem Soc 125 
(2003) 14093-14102. 
[166] S. Chouchane, S. Girotto, S. Kapetanaki, J.P.M. Schelvis, S.W. Yu, R.S. Magliozzo, 
Analysis of heme structural heterogeneity in Mycobacterium tuberculosis catalase-
peroxidase (KatG), J Biol Chem 278 (2003) 8154-8162. 
[167] D.A. Svistunenko, J.A. Worrall, S.B. Chugh, S.C. Haigh, R.A. Ghiladi, P. Nicholls, 
Ferric haem forms of Mycobacterium tuberculosis catalase-peroxidase probed by EPR 
spectroscopy: Their stability and interplay with pH, Biochimie 94 (2012) 1274-1280. 
[168] G.R. Sutherland, S.D. Aust, The effects of calcium on the thermal stability and activity of 
manganese peroxidase, Arch Biochem Biophys 332 (1996) 128-134. 
 194 
[169] M. Sundaramoorthy, K. Kishi, M.H. Gold, T.L. Poulos, The crystal structure of 
manganese peroxidase from Phanerochaete chrysosporium at 2.06-A resolution, J Biol 
Chem 269 (1994) 32759-32767. 
[170] Y. Nishizawa, Z. Nishio, K. Nakazono, M. Soma, E. Nakajima, M. Ugaki, T. Hibi, 
Enhanced resistance to blast (Magnaporthe grisea) in transgenic Japonica rice by 
constitutive expression of rice chitinase, Theoretical and Applied Genetics 99 (1999) 
383-390. 
[171] K. Soderhall, Special issue: Invertebrate immunity, Developmental and Comparative 
Immunology 23 (1999) 263-266. 
[172] Y. Wang, D. Goodwin, Integral role of the I'-helix in the function of the "inactive" C-
terminal domain of catalase-peroxidase (KatG). Biochimica et Biophysica Acta, 2012, In 
Press 
[173] A. Fersht, Structure and Mechanism in Protein Science: a Guide to Enzyme Catalysis and 
Proyein Folding. W.H. Freeman and Company, New York, 1999 
[174] J. Peterson, C. Vibat, R.B. Gennis, Identification of the axial heme ligands of cytochrome 
b556 in succinate: ubiquinone oxidoreductase from Escherichia coli, FEBS Lett, 355 
(1994) 155-156. 
[175] E.A. Berry, F.A. Walker, Bis-histidine-coordinated hemes in four-helix bundles: how the 
geometry of the bundle controls the axial imidazole plane orientations in transmembrane 
cytochromes of mitochondrial complexes II and III and related proteins, J. Biol. Inorg. 
Chem. 13 (2008) 481-498. 
[176] G. Zoppellaro, K.L. Bren, A.A. Ensign, E. Harbitz, R. Kaur, H.P. Hersleth, U. Ryde, L. 
Hederstedt, K.K. Andersson, Review: studies of ferric heme proteins with highly 
 195 
anisotropic/highly axial low spin (S = 1/2) electron paramagnetic resonance signals with 
bis-histidine and histidine-methionine axial iron coordination, Biopolymers 91 (2009) 
1064-1082. 
[177] G.R. Sutherland, L.S. Zapanta, M. Tien, S.D. Aust, Role of calcium in maintaining the 
heme environment of manganese peroxidase, Biochemistry-Us 36 (1997) 3654-3662. 
[178] G. Nie, S.D. Aust, Spectral changes of lignin peroxidase during reversible inactivation, 
Biochemistry-Us 36 (1997) 5113-5119. 
[179] G. Nie, S.D. Aust, Effect of calcium on the reversible thermal inactivation of lignin 
peroxidase, Arch Biochem Biophys 337 (1997) 225-231. 
[180] D.J. Schuller, N. Ban, R.B. Huystee, A. McPherson, T.L. Poulos, The crystal structure of 
peanut peroxidase, Structure 4 (1996) 311-321. 
[181] R. Moore, Toward the Understanding of Complex Biochemical Systems: the Significance 
of Global Protein Structure and Thorough Parametric Analysis 2008, Dissertation, 
Auburn University. 
[182] A. Hillar, B. Peters, R. Pauls, A. Loboda, H. Zhang, A.G. Mauk, P.C. Loewen, 
Modulation of the activities of catalase-peroxidase HPI of Escherichia coli by site-
directed mutagenesis, Biochemistry-Us 39 (2000) 5868-5875. 
[183] G. Regelsberger, C. Jakopitsch, F. Ruker, D. Krois, G.A. Peschek, C. Obinger, Effect of 
distal cavity mutations on the formation of compound I in catalase-peroxidases, J Biol 
Chem 275 (2000) 22854-22861. 
[184] J.E. Erman, L.B. Vitello, M.A. Miller, A. Shaw, K.A. Brown, J. Kraut, Histidine-52 is a 
critical residue for rapid formation of cytochrome-c peroxidase compound-I, 
Biochemistry-Us 32 (1993) 9798-9806. 
 196 
[185] L.B. Vitello, J.E. Erman, M.A. Miller, J. Wang, J. Kraut, Effect of arginine-48 
replacement on the reaction between cytochrome c peroxidase and hydrogen peroxide, 
Biochemistry-Us 32 (1993) 9807-9818. 
[186] D.B. Goodin, D.E. McRee, The Asp-His-Fe triad of cytochrome c peroxidase controls the 
reduction potential, electronic structure, and coupling of the tryptophan free radical to the 
heme, Biochemistry-Us 32 (1993) 3313-3324. 
[187] K.G. Welinder, B. Bjornholm, H.B. Dunford, Functions of electrostatic potentials and 
conserved distal and proximal His-Asp H-bonding networks in haem peroxidases, 
Biochem Soc Trans 23 (1995) 257-262. 
[188] J.D. Satterlee, S.L. Alam, J.M. Mauro, J.E. Erman, T.L. Poulos, The effect of the Asn82--
>Asp mutation in yeast cytochrome c peroxidase studied by proton NMR spectroscopy, 
Eur J Biochem / FEBS 224 (1994) 81-87. 
[189] S.F. DeLauder, J.M. Mauro, T.L. Poulos, J.C. Williams, F.P. Schwarz, Thermodynamics 
of hydrogen cyanide and hydrogen fluoride binding to cytochrome c peroxidase and its 
Asn-82-->Asp mutant, Biochem J 302 (1994) 437-442. 
[190] M. Tanaka, S. Nagano, K. Ishimori, I. Morishima, Hydrogen bond network in the distal 
site of peroxidases: spectroscopic properties of Asn70 --> Asp horseradish peroxidase 
mutant, Biochemistry-Us 36 (1997) 9791-9798. 
[191] R.L. Moore, C.O. Cook, R. Williams, D.C. Goodwin, Substitution of strictly conserved 
Y111 in catalase-peroxidases: Impact of remote interdomain contacts on active site 
structure and catalytic performance, J Inorg Biochem. 102 (2008) 1819-1824. 
 197 
[192] J.E. Erman, L.B. Vitello, M.A. Miller, A. Shaw, K.A. Brown, J. Kraut, Histidine 52 is a 
critical residue for rapid formation of cytochrome c peroxidase compound I, 
Biochemistry-Us 32 (1993) 9798-9806. 
[193] N.L. Wengenack, M.P. Jensen, F. Rusnak, M.K. Stern, Mycobacterium tuberculosis 
KatG is a peroxynitritase, Biochem Biophys Res Commun, 256 (1999) 485-487. 
[194] T. Sagawa, T. Yomo, S. Tsugawa, Y. Takeda, H. Ihara, E.K. Mitamura, I. Urabe, K. 
Ohkubo, A novel versatility of catalase I as a dioxygenase for indole-ring-opening 
dioxygenation, Chem Lett  (1999) 339-340. 
[195] P.C. Loewen (1997) Bacterial catalase. In Oxidative stress and the molecular biology of 
antioxidant defenses. pp 273-308, Cold Spring Harbor Laboratory Press, Plainview, N.Y 
[196] E. Santoni, C. Jakopitsch, C. Obinger, G. Smulevich, Comparison between catalase-
peroxidase and cytochrome c peroxidase. The role of the hydrogen-bond networks for 
protein stability and catalysis, Biochemistry-Us 43 (2004) 5792-5802. 
[197] C. Jakopitsch, A. Ivancich, F. Schmuckenschlager, A. Wanasinghe, G. Poltl, P.G. 
Furtmuller, F. Ruker, C. Obinger, Influence of the unusual covalent adduct on the 
kinetics and formation of radical intermediates in synechocystis catalase peroxidase: a 
stopped-flow and EPR characterization of the MET275, TYR249, and ARG439 variants, 
J Biol Chem  279 (2004) 46082-46095. 
[198] S.M. Kapetanaki, X.B. Zhao, S.W. Yu, R.S. Magliozzo, J.P.M. Schelvis, Modification of 
the active site of Mycobacterium tuberculosis KatG after disruption of the Met-Tyr-Trp 
cross-linked adduct, J Inorg Biochem 101 (2007) 422-433. 
[199]  C.L. Varnado,  Enhancing expression of recombinant hemoproteins: progress toward 
understanding structure/function and  therapeutic application  (2006) Dissertation, 
 198 
Auburn University.