Selective tuning of activity in a multifunctional enzyme as revealed in the F21W mutant of dehaloperoxidase B from Amphitrite ornata

Article abstract of DOI:10.1007/s00775-017-1520-x

Possessing both peroxidase and peroxygenase activities with a broad substrate profile that includes phenols, indoles, and pyrroles, the enzyme dehaloperoxidase (DHP) from Amphitrite ornata is a multifunctional catalytic hemoglobin that challenges many of the assumptions behind the well-established structure–function paradigm in hemoproteins. While previous studies have demonstrated that the F21W variant leads to attenuated peroxidase activity in DHP, here we have studied the impact of this mutation on peroxygenase activity to determine if it is possible to selectively tune DHP to favor one function over another. Biochemical assays with DHP B (F21W) revealed minimal decreases in peroxygenase activity of 1.2–2.1-fold as measured by 4-nitrophenol or 5-Br-indole substrate conversion, whereas the peroxidase activity catalytic efficiency for 2,4,6-trichlorophenol (TCP) was more than sevenfold decreased. Binding studies showed a 20-fold weaker affinity for 5-bromoindole (K_d?=?2960?±?940?μM) in DHP B (F21W) compared to WT DHP B. Stopped-flow UV/visible studies and isotope labeling experiments together suggest that the F21W mutation neither significantly changes the nature of the catalytic intermediates, nor alters the mechanisms that have been established for peroxidase and peroxygenase activities in DHP. The X-ray crystal structure (1.96??; PDB 5VLX) of DHP B (F21W) revealed that the tryptophan blocks one of the two identified TCP binding sites, specifically TCP_interior, suggesting that the other site, TCP_exterior, remains viable for binding peroxygenase substrates. Taken together, these studies demonstrate that blocking the TCP_interior binding site in DHP selectively favors peroxygenase activity at the expense of its peroxidase activity. Graphical abstract: [Figure not available: see fulltext.].

Full text of DOI:10.1007/s00775-017-1520-x

JBIC Journal of Biological Inorganic Chemistry
https://doi.org/10.1007/s00775-017-1520-x
ORIGINAL PAPER
Selective tuning of activity in a multifunctional enzyme as revealed
in the F21W mutant of dehaloperoxidase B from Amphitrite ornata
Leiah M. Carey · Kyung Beom Kim1,2 · Nikolette L. McCombs · Paul Swartz · Cheal Kim · Reza A. Ghiladi
1
1
3
2
1
Received: 28 September 2017 / Accepted: 17 November 2017
©
SBIC 2017
Abstract
Possessing both peroxidase and peroxygenase activities with a broad substrate profle that includes phenols, indoles, and
pyrroles, the enzyme dehaloperoxidase (DHP) ꢀrom Amphitrite ornata is a multiꢀunctional catalytic hemoglobin that chal-
lenges many oꢀ the assumptions behind the well-established structure–ꢀunction paradigm in hemoproteins. While previous
studies have demonstrated that the F21W variant leads to attenuated peroxidase activity in DHP, here we have studied the
impact oꢀ this mutation on peroxygenase activity to determine iꢀ it is possible to selectively tune DHP to ꢀavor one ꢀunction
over another. Biochemical assays with DHP B (F21W) revealed minimal decreases in peroxygenase activity oꢀ 1.2–2.1-
ꢀ
old as measured by 4-nitrophenol or 5-Br-indole substrate conversion, whereas the peroxidase activity catalytic eꢁciency
ꢀ
or 2,4,6-trichlorophenol (TCP) was more than sevenꢀold decreased. Binding studies showed a 20-ꢀold weaker aꢁnity ꢀor
5
-bromoindole (K = 2960 ± 940 μM) in DHP B (F21W) compared to WT DHP B. Stopped-ꢂow UV/visible studies and
d
isotope labeling experiments together suggest that the F21W mutation neither signifcantly changes the nature oꢀ the catalytic
intermediates, nor alters the mechanisms that have been established ꢀor peroxidase and peroxygenase activities in DHP. The
X-ray crystal structure (1.96 Å; PDB 5VLX) oꢀ DHP B (F21W) revealed that the tryptophan blocks one oꢀ the two identifed
TCP binding sites, specifcally TCP
, suggesting that the other site, TCP
, remains viable ꢀor binding peroxygenase
binding site in DHP selectively ꢀavors
interior
exterior
substrates. Taken together, these studies demonstrate that blocking the TCP
peroxygenase activity at the expense oꢀ its peroxidase activity.
Graphical abstract
interior
Peroxygenase Acꢀve Peroxidase Inhibited
Trp21
DHP B (F21W)
Keywords Globin · Peroxidase · Peroxygenase · Structure–ꢀunction relationship
Introduction
Electronic supplementary material The online version oꢀ this
article (https://doi.org/10.1007/s00775-017-1520-x) contains
supplementary material, which is available to authorized users.
The coelomic hemoglobin oꢀ the sediment-dwelling
marine worm Amphitrite ornata ꢀunctions as a naturally
occurring dehalogenation enzyme that has been termed
Extended author inꢀormation available on the last page oꢀ the article
Vol.:(0
1
123456789)
3

JBIC Journal of Biological Inorganic Chemistry
TCPexterior
TCPinterior
observed to bind in a single site and in a similar orienta-
tion as TCP
. The observation oꢀ multiple binding sites
interior
across the entire ꢀace oꢀ the distal heme cavity is consistent
with DHP being a multiꢀunctional catalytic globin. More
importantly, one can evaluate the orientation oꢀ binding ꢀor
these substrates in Fig. 1 against the activities that each is
subjected to: the unhindered (at the C2 and C3 positions)
nitrophenol substrate is susceptible to DHP peroxygenase
activity, possibly via an electrophilic addition by the ꢀerryl
intermediate, leading to incorporation oꢀ the O atom ꢀrom
hydrogen peroxide [4]. By contrast, the comparable posi-
tions in TBP are sterically hindered by the halogen substitu-
ents, precluding O atom transꢀer, and leading to substrate
oxidation via a peroxidase mechanism [27]. TCP binds as a
Heme δ edge
Fig. 1 Superposition oꢀ substrate binding sites within the distal
pocket oꢀ DHP as viewed ꢀrom the heme γ edge: TBP (yellow, PDB
4
FH7 [27]), 4-NP (purple, PDB 5CHQ [4]), 4-BP (green, PDB 3LB2
[
26]), and both internal (TCP
) and external (TCP
) conꢀor-
interior
exterior
mations oꢀ TCP (cyan, PDB 4KN3 [24])
peroxidase substrate (i.e., TCP
) in a conꢀormation that
interior
closely resembles that observed ꢀor TBP, but selꢀ-inhibition
dehaloperoxidase (DHP) [1, 2]. In addition to being able
to oxidize/degrade 7 EPA priority pollutants oꢀ phenolic
structure (phenol, 2-chlorophenol, 2,4-dichlorophenol,
at high concentrations has also been observed [24], and has
been attributed to the TCP
site as this conꢀormation
exterior
orients the para-halogen above the heme iron and to the
2
2
,4,6-trichlorophenol, 2-nitrophenol, 4-nitrophenol, and
,4-dinitrophenol) [3, 4], DHP more broadly has shown
detriment oꢀ H O -binding. Although currently there is no
2
2
indole-bound structure oꢀ DHP, both computational studies
oꢀ haloindole binding [5] and X-ray structures with bound
azoles [23] (as models ꢀor indole) strongly suggest that
indoles bind in an orientation relative to the ꢀerryl heme that
ꢀacilitates O atom transꢀer via a peroxygenase mechanism.
Following the identifcation oꢀ the two TCP binding
sites in DHP, Dawson and co-workers rationally designed
the F21W mutation in DHP A to block one oꢀ those sites,
activity against halogenated indoles [5], pyrroles [6], and
guaiacols. Mechanistic studies [7–18] have shown that DHP
appears to ꢀunction via a Poulos–Kraut type mechanism
[
19] in which H O reacts with a ꢀerric heme to ꢀorm DHP
2 2
compound I [13, 16], the iron(IV)-oxo (ꢀerryl) porphyrin
π-cation radical species that is common to both the peroxi-
dase [20] and P450 cycles [21]. Accordingly, DHP acts on
the aꢀorementioned substrates with peroxidase [2, 22], per-
oxygenase [5, 6], and oxidase [5] mechanisms. Moreover,
DHP is known to bind a number oꢀ persistent organic pol-
lutants (POPs) belonging to the azole class oꢀ compounds
TCP
[24]. Modeling revealed that the potential W21
interior
rotamers oꢀ the mutant either blocked the TCP
site, or
interior
were disallowed because oꢀ steric hindrance with the pro-
tein. The consequence oꢀ the F21W mutation was a 15-ꢀold
TCP
[
23], including benzotriazole, benzimidazole, indazole, and
reduction in catalytic eꢁciency (k /K ) vs WT DHP
cat
M
imidazole that have widespread agricultural and industrial
uses. Thus, given the broad substrate scope and multiple
catalytic activities exhibited by DHP, one oꢀ the outstand-
ing questions is how substrate binding is related to enzyme
activity in this multiꢀunctional and promiscuous catalytic
globin, and whether DHP can be tuned to ꢀavor one activity
over another.
A ꢀor peroxidase activity with TCP as the substrate, thus
confrming their hypothesis that mutagenesis oꢀ F21 to the
larger Trp residue would signifcantly impair peroxidase
activity. At that time, however, the peroxygenase [4–6]
activity oꢀ DHP was not known, and the consequence oꢀ
the F21W mutation on that activity was unexplored. Given
the recent identifcation oꢀ the 4-nitrophenol peroxyge-
nase substrate binding site [4] that does not overlap sig-
One oꢀ the key ꢀeatures oꢀ DHP that sets it apart ꢀrom tra-
ditionally studied globins (e.g., myoglobin and hemoglobin),
and even many heme peroxidases, is that it possesses a distal
small molecule binding site that can accommodate a variety
oꢀ diꢃerent substrates (and inhibitors) [1, 2, 24, 25], exam-
ples oꢀ which are shown in Fig. 1 and include the peroxyge-
nase substrate 4-nitrophenol (4-NP, purple, PDB 5CHQ [4]),
inhibitor 4-bromophenol (4-BP, PDB 3LB2 [26]), and the
peroxidase substrates 2,4,6-tribromophenol (TBP, yellow,
PDB 4FH7 [27]) and 2,4,6-trichlorophenol (TCP, cyan, PDB
nifcantly with TCP
, and the unknown consequence
interior
oꢀ the F21W mutation on peroxygenase activity, we were
motivated to explore this mutation as a possible means
oꢀ selectively tuning the activity profle oꢀ DHP. To that
end, we present here a study oꢀ the DHP B (F21W) mutant
on peroxygenase activity, including biochemical assays,
stopped-ꢂow UV–visible spectroscopic characterization oꢀ
the activated enzyme, as well as an X-ray crystallographic
study oꢀ the mutant. The results demonstrate the ability to
ꢀavor peroxygenase activity over peroxidase activity in a
rationally designed mutant oꢀ a multiꢀunctional catalytic
globin.
4
KN3 [24]). Two substrate binding sites were determined
ꢀ
(
or TCP [TCP (near the heme α edge) and TCP
interior
exterior
near the heme γ edge and heme cavity)], while TBP was
1
3

JBIC Journal of Biological Inorganic Chemistry
Experimental
Protein crystallization and X‑ray diꢀraction studies
Materials
Non-His tagged DHP B (F21W) was overexpressed and
purifed ꢀollowing the literature protocol [4]. Crystals were
obtained through the hanging-drop vapor diꢃusion method.
DHPB (F21W), in 20 mM Na cacodylate buꢃer pH 6.4, was
concentrated to 12 mg/mL, and the crystals were grown ꢀrom
mother liquor solutions oꢀ 28–32% PEG 4000 and 0.2 M
ammonium sulꢀate at pH 6.4, equilibrated against identi-
cal reservoir solutions. Protein-to-mother liquor ratios var-
ied between 1:1, 1.33:1, 1.66:1 and 2:1. At 4 °C, crystals
grew ꢀrom each condition aꢀter 3 days. The crystals were
cryo-protected by brieꢂy dipping them in reservoir solu-
tion enhanced with 20% glycerol and then ꢀlash ꢀrozen
Unless otherwise speciꢀied, all chemicals were oꢀ rea-
gent grade, purchased ꢀrom VWR or Fisher Scientiꢀic,
1
8
18
and used without ꢀurther puriꢀication. H O and H O
2
2
2
(
90 and 98% oxygen atom enriched, respectively) were
obtained ꢀrom ICON Isotopes (Summit, NJ, USA). DHP
was expressed, puriꢀied and obtained in the ꢀerric oxida-
tion state as previously reported [9]. The enzyme concen-
tration was determined spectrophotometrically using the
−
1
−1
molar absorptivity coeꢀꢀicient (ε = 116,400 M cm
4
06
[
9]). Stock solutions (2 mM) oꢀ trichlorophenol (TCP)
were prepared in 100 mM potassium phosphate (KP )
in liquid N . Data were collected at the Biological X-ray
i
2
buꢀꢀer (pH 7) and stored at − 80 °C. TCP concentration
Facility at NCSU on a Rigaku RuH3R with rotating copper
anode equipped with a MAR345 CCD image plate detector
and Oxꢀord cryojet set at 100 K, utilizing a wavelength oꢀ
1.54 Å. All data were scaled and integrated using HKL2000
[29], molecular replacement was perꢀormed with Phaser-MR
[30] ꢀrom the PHENIX [31] suite oꢀ programs using the
3IXF [18] monomer as the search model, model building
and manual placement oꢀ waters utilized COOT [32] and
refnement was carried out using phenix.refne [33].
and lack oꢀ degradation was monitored by measuring its
−
1
−1
absorbance at 312 nm (ε = 3752 M cm [8]). Stock
solutions (10 mM) oꢀ 5-bromoindole and 7-bromoindole
were prepared in MeOH, stored at − 80 °C in glass vials,
and screened by HPLC ꢀor degradation prior to use [5].
Aliquots were stored on ice during use. Stock solutions
(
5 mM) oꢀ 4-nitrophenol (4-NP) were prepared ꢀresh in
00 mM KPi (pH 7) when needed. Solutions oꢀ H O
1
2
2
were prepared daily, and stored on ice while protected
ꢀ
rom light until needed. The concentration oꢀ H O was
Peroxidase studies
2
2
−
1
−1
determined spectrophotometrically (ε = 43.6 M cm
2
40
[
28]).
The hydrogen peroxide-dependent oxidative dehalogenation
oꢀ 2,4,6-trichlorophenol (TCP) to 2,6-dichloro-1,4-benzo-
quinone (DCQ) as catalyzed by DHP was measured using
a Cary 50 UV–Vis spectrophotometer. The reactions were
Construction of mutant DHP plasmid
perꢀormed in triplicate at pH 7 in 100 mM KP at 25 °C, with
i
Site-directed mutagenesis was perꢀormed using the
Quikchange II site-directed mutagenesis kit (Agilent
Technologies). Mutagenesis [melt (95 °C, 60 s), anneal
a 1 mL total reaction volume. Buꢃered solutions oꢀ DHP
(1.25 µM) and TCP (10–800 µM) were premixed and the
reaction was initiated by the addition oꢀ H O (80 µM). The
2
2
−
1
−1
(
55 °C, 50 s), and extension (68 °C, 360 s)] was per-
ormed ꢀor 16 cycles. Oligonucleotides were synthetized
change in absorbance at 272 nm (DCQ, ε = 14 mM cm
ꢀ
[7]) was measured using Cary WinUV Kinetics soꢀtware
over a 30 s time ꢀrame with a rate oꢀ 1 scan per second.
The data were ft to the standard Michaelis–Menten kinetics
model using the method oꢀ initial rates in the Graft soꢀt-
ware package (Erithacus Soꢀtware). The apparent kinetics
parameters K and k resulted ꢀrom the optimization oꢀ
by Integrated DNA Technologies (IDT) (sense 5′-ACC
TAT GCA CAG GAC ATT TGG CTC GCA TTT TTG
AAT AGG-3′; anti-sense 5′-CCT ATT CAA AAA TGC
GAG CCA AAT GTC CTG TGC ATA GGT-3′). The
plasmid encoding a N-terminal poly-His tag wild-type
DHP B was used as a template [9]. For crystallization, the
mutant was obtained ꢀrom a plasmid-encoding wild-type
DHP B lacking the N-terminal poly-His tag [18]. The
modiꢀied plasmids were transꢀormed into E. coli BL21-
Gold(DE3) competent cells (Agilent Technologies) and
selected based on survival on LB-agar-ampicillin (100 µg/
mL) plates. The plasmids were extracted using the spin
column plasmid DNA kit (Bio Basic), and the desired
mutation and lack oꢀ deleterious secondary ones were
conꢀirmed by sequencing (Genewiz).
M
cat
the ftting procedure.
Peroxygenase studies
The percent conversions oꢀ 5-Br-indole, 7-Br-indole and
4-nitrophenol (4-NP) as catalyzed by DHP in the pres-
ence oꢀ H O were analyzed by HPLC using a Waters 2796
2
2
Bioseparations Module coupled with a Waters 2996 Photo-
diode Array Detector and equipped with an ODS Hypersil
C
column (Thermo-Scientifc, 150 mm × 4.6 mm, 5 µm
8
1
particle size). The reactions were perꢀormed in triplicate in
1
3

JBIC Journal of Biological Inorganic Chemistry
1
00 mM KP (pH 7) at 25 °C. Indole reactions also contained
Results
i
5
% MeOH (v/v). Buꢃered solutions oꢀ substrate and DHP
were premixed, and the peroxygenase reaction was initiated
with addition oꢀ H O . Final concentrations were 10 µM
Overexpression, puriꢂcation and characterization
of DHP B
2
2
DHP, 500 µM substrate and 500 µM H O in 200 µL total
2
2
volume. Aꢀter 5 min oꢀ incubation at 25 °C, the reaction
Recombinant DHP B (F21W) was obtained by expres-
sion in E. coli and purifed as previously described [4, 9,
18]. The monomeric molecular weight oꢀ DHP B (F21W)
was determined by electrospray ionization MS to be
16,313.11 g/mol, in good agreement with the theoretical
expected value (16,313.42 g/mol). DHP B (F21W) was
initially isolated as a mixture oꢀ the ꢀerric and oxyꢀerrous
ꢀorms, as were WT isoenzymes A and B [8, 9], and subse-
quent treatment with an excess oꢀ potassium ꢀerricyanide
permitted the isolation oꢀ the ꢀerric ꢀorm. The electronic
absorption spectrum oꢀ ꢀerric DHP B (F21W) exhibited
was quenched with excess catalase and then diluted 1:10
with 1800 µL oꢀ 100 mM KP (pH 7). Diluted samples were
i
subjected to HPLC analysis (solvent A—H O containing
2
0
.1% triꢂuoroacetic acid: solvent B—HPLC-grade MeCN
containing 0.1% triꢂuoroacetic acid). The elution method
was as ꢀollows: 1.5 mL/min oꢀ 95:5 (A:B) to 5:95 using a lin-
ear gradient over 10 min; 5:95 isocratic ꢀor 2 min; 5:95–95:5
using a linear gradient over 1 min, and then isocratic ꢀor
4
min. Analysis was perꢀormed using the Empower soꢀtware
package (Waters Corp.).
ꢀ
eatures typical oꢀ a high-spin ꢀerric heme [UV–visible:
5
‑Br‑indole binding studies
407 (Soret), 504, 634 nm; Fig. 2] that were similar to
those observed previously ꢀor WT DHP B [UV–visible:
407 (Soret), 508, 633 nm] [9]. The optical purity ratio
As adapted ꢀrom previously published protocols [5, 34], the
5
-Br-indole dissociation constant (K ) was determined in
(Reinheitzahl or Rz, defned as A
/A ) ꢀor DHP B
Soret 280
d
triplicate ꢀor DHP in 100 mM KP (pH 7) containing 10%
(F21W), was ꢀound to be 3.63, lower than the literature
value ꢀor WT DHP B oꢀ 4.1 [9, 13]. This was expected,
however, as the F21W mutation increases the calculated
i
MeOH at 25 °C using a Cary 50 UV–Vis spectrophotom-
eter. A stock solution oꢀ 10 mM 5-Br-indole in MeOH was
prepared in a glass vial. The UV–Vis spectrophotometer was
−
1
−1
A
molar absorptivity coeꢁcient to 16,960 M cm in
80
2
−
1
−1
reꢀerenced with 10 µM ꢀerric DHP in 100 mM KP (pH 7)
DHP B (F21W) ꢀrom 11,460 M cm in WT DHP B.
i
containing 10% MeOH (v/v). The total volume oꢀ the sam-
ples was 200 µL with fnal concentrations oꢀ 10 µM ꢀerric
DHP and 10–130 equivalents 5-Br-indole while maintain-
ing 10% MeOH (v/v). Perturbations in the absorbance oꢀ
the Soret band (ΔAbs) were recorded ꢀor each 5-Br-indole
concentration (Figure S1). Analysis by nonlinear regression
using the GraFit soꢀtware package (Erithacus Soꢀtware Ltd.)
The A /A (1.87) ratio ꢀor DHP B (F21W) was also
Soret 380
similar to that ꢀor WT DHP B (A
/A = 1.88; A
/
614
Soret 380
A
= 1.14) [9, 13]. However, the A /A (0.93) ratio
614 645
6
45
was lower in the F21W mutant, suggesting a slightly lower
population oꢀ six-coordinate high-spin heme in the mutant
when compared to its wild-type analog. Sodium dithion-
ite reduction oꢀ the ꢀerric enzyme in the presence oꢀ O
2
provided a calculated A , which was used to calculate α
yielded oxyꢀerrous DHP B (F21W), with spectral ꢀeatures
max
(
ΔA/ΔA ) ꢀor the average ΔA ꢀor each concentration. A
[UV–visible: 418 (Soret), 544, 578 nm] that were again
max
nonlinear regression plot provided the reported K values.
d
Stopped‑ꢁow UV–visible studies
4
07
ferric DHP B (F21W)
1
1
20
00
4
18
oxyferrous DHP B (F21W)
Optical spectra were recorded using a Bio-Logic SFM-400
triple-mixing stopped-ꢂow instrument coupled to a rapid
scanning diode array UV–visible spectrophotometer. The
reactions were conducted at room temperature, and all solu-
544 578
16
12
8
504
8
6
4
2
0
0
0
0
0
634
tions were prepared in 100 mM KP (pH 7). Data were col-
4
i
lected (900 scans total) over a three-time domain regime
0
4
50
500
550
600
650
700
(
2.5, 25, and 250 ms; 300 scans each) using the Bio Kinet32
soꢀtware package (Bio-Logic). All data were evaluated using
the Specft Global Analysis System soꢀtware package (Spec-
trum Soꢀtware Associates) and ft to exponential ꢀunctions
as two-step/three-species irreversible mechanisms. Experi-
ments were perꢀormed in single-mixing mode where DHP B
3
00 350 400 450 500 550 600 650 700 750 800
Wavelength (nm)
(
F21W) at a fnal concentration oꢀ 10 μM was reacted with
.5–25 equivalents oꢀ H O .
Fig. 2 UV–visible spectra oꢀ ꢀerric (red) and oxyꢀerrous (black) DHP
2
B (F21W) in 100 mM KP (pH 7)
2
2
i
1
3

JBIC Journal of Biological Inorganic Chemistry
consistent with the six-coordinate low-spin oxyꢀerrous
higher by 1.7-ꢀold (34.0 ± 1.6 vs 20.3 ± 1.7% ꢀor WT DHP
A [5]), while the 4-NP reactivity is approximately equivalent
ꢀorms oꢀ the wild-type isoenzymes [8, 9].
(
19.1 ± 0.3 vs 21.6 ± 0.1% ꢀor WT DHP A [4]). As was the
Activity assays
case ꢀor peroxidase activity, inclusion oꢀ 500 μM 4-bromo-
phenol as an inhibitor attenuated 5-Br-indole substrate con-
version to 4.6 ± 1.4%.
The peroxidase activity oꢀ DHP ꢀor the oxidative dehalo-
genation oꢀ 2,4,6-trichlorophenol (TCP) to 2,6-dichloro-
1
,4-benzoquinone (DCQ) was monitored spectrophotometri-
Isotopically labeled oxygen studies
cally in a (TCP)-dependent manner at a fxed concentration
oꢀ H O . When compared with the kinetic parameters oꢀ WT
1
8
18
Studies employing labeled H O and H O (90 and 98%
2
2
2
2
2
−
1
DHP B (Table 1; k = 1.90 ± 0.20 s ; K = 90 ± 20 µM;
oxygen atom enriched, respectively) were perꢀormed with
5-Br-indole and subsequently analyzed by LC–MS to con-
frm hydrogen peroxide as the source oꢀ the O atom incor-
porated in the peroxygenase studies. The background-sub-
tracted total ion chromatograms (TICs) are shown in Fig. 3
ꢀor the 5-Br-2-oxindole product. Reactions perꢀormed in the
cat
M
−1 −1
k /K = 21 ± 5 mM s ), DHP B (F21W) exhibited a 1.5-
cat
M
−
1
ꢀ
old decrease in catalytic turnover (kcat = 1.29 ± 0.07 s ), a
4
.9-ꢀold increase in K (440 ± 50 µM), and a 7.2-ꢀold lower
M
−
1 −1
catalytic eꢁciency (k /K = 2.9 ± 0.3 mM s ). These
cat
M
results demonstrate that the F21W mutation signifcantly
aꢃects the peroxidase activity oꢀ DHP B, but the eꢃect is
less pronounced that observed ꢀor DHP A (F21W) when
1
8
presence oꢀ H O and unlabeled H O show 5-Br-2-oxindole
2
2
2
(m/z 212, 214; Fig. 3a) with product m/z values identical to
compared to WT DHP A (3.3-ꢀold lower k , 4.6-ꢀold higher
those obtained ꢀrom the unlabeled H O/H O reaction. Grat-
cat
2
2
2
1
8
K , and 15-ꢀold lower k /K ) [24]. When 4-bromophenol
iꢀyingly, the reactions perꢀormed in the presence oꢀ H O in
M
cat
M
2 2
(
K = 1.15 mM [26]), the known inhibitor oꢀ DHP peroxi-
unlabeled H O exhibited the characteristic m/z shiꢀt oꢀ +2
2
d
1
8
dase and peroxygenase activities [5, 9], was included in the
reaction at 500 μM, the rate oꢀ TCP oxidation as catalyzed
by the F21W mutant was reduced by 45%.
amu (Fig. 3b), denoting O-label incorporation. Similarly,
the 5-Br-3-oxindolenine (m/z 210, 212) product only showed
1
8
a shiꢀt oꢀ + 2 amu in the presence oꢀ labeled H O (data not
2
2
The peroxygenase activity oꢀ DHP was monitored by
HPLC ꢀor the H O -dependent oxidation oꢀ 5-Br-indole,
shown). Taken together, the labeling studies confrmed that
the oxygen atom incorporated into the products was derived
solely ꢀrom H O , thus demonstrating that the F21W muta-
2
2
7
-Br-indole, and 4-nitrophenol (4-NP) when initiated ꢀrom
2
2
the ꢀerric state at pH 7. Overall, the eꢃect oꢀ the F21W muta-
tion on substrate percent conversion was limited: the DHP
B (F21W) mutant exhibited a decrease oꢀ: (i) 1.4-ꢀold ꢀor
tion does not alter the reported peroxygenase chemistry.
Stopped‑ꢁow UV–visible spectroscopy
5
-Br-indole (34.0 ± 1.6 vs 48.1 ± 2.3% ꢀor WT DHP B [5]),
(
ii) 1.2-ꢀold ꢀor 7-Br-indole (39.4 ± 1.2 vs 46.1 ± 1.7% ꢀor
WT DHP B [5]), and (iii) 2.1-ꢀold ꢀor 4-NP (19.1 ± 0.3 vs
9.4 ± 0.7% ꢀor WT DHP B [4]). As peroxygenase activity
was not known at the time when Dawson and co-workers
24] frst described the eꢃects oꢀ the F21W mutation on the
catalytic (peroxidase) activity oꢀ DHP, no data ꢀor DHP A
F21W) are available. However, when compared with WT
DHP A, the reactivity oꢀ DHP B (F21W) with 5-Br-indole is
To assess the ability oꢀ DHP B (F21W) to ꢀorm compound
ES, the catalytically active two-electron-oxidized state
IV
3
containing both a ꢀerryl center [Fe =O] and a tyrosyl
radical, single-mixing stopped-ꢂow UV–visible spectro-
scopic methods were employed as previously described [8,
9]. Using the 10-eq. H O reaction as representative data
[
2
2
(
(Fig. 4), the ꢀollowing observations were made: an SVD
analysis employing a simple three component, irreversible
Table 1 Kinetic parameters ꢀor the oxidation oꢀ TCP, and percent substrate conversion ꢀor the oxygenation oꢀ 5-Br-indole, 7-Br-indole and 4-NP
Enzyme
Peroxidasea
Peroxygenasea,b
Reꢀs.
TCP
kcat (s−1
kcat/KM (mM−1 s−1
K
(μM)
)
)
5-Br-indole
7-Br-indole
4-nitrophenol
M
c
c
DHP B (F21W)
WT DHP B
DHP A (F21W)
WT DHP A
440 ± 50
90 ± 20
2260 ± 512
495 ± 62
1.29 ± 0.07
1.90 ± 0.20
0.32 ± 0.05
1.05 ± 0.07
2.9 ± 0.3
21 ± 5
0.14 ± 0.03
2.1 ± 0.3
34.0 ± 1.6
48.1 ± 2.3
n/a
39.4 ± 1.0
46.1 ± 1.7
n/a
19.1 ± 0.3
39.4 ± 0.7
n/a
[24]
[4, 5, 24]
20.3 ± 1.7
n/a
21.6 ± 0.1
n/a not available
a
Reactions were perꢀormed with ꢀerric DHP in 100 mM KP buꢃer (pH 7) at 25 °C
Values represent percent (%) substrate conversion
This work
i
b
c
1
3

JBIC Journal of Biological Inorganic Chemistry
Fig. 3 ESI-MS total ion chro-
matograms obtained ꢀor the
reaction product 5-Br-2-oxin-
A
H218O/H 16O
2
B
H216O/H 18O
2
2
2
2
0
5
4
3
2
1
0
2
12
214
216
1
8
16
16
dole (A: H O, H O ; B: H O,
214
2
2
2
2
1
8
H O ). Reaction conditions:
15
2
2
[
5-Br-indole] = [H O ] = 500
Br
Br
2
2
μM, [enzyme] = 10 μM,
00 mM KP (pH 7), 25°C
O
O
1
0
5
0
N
H
N
H
1
i
2
10 212 214 216 218
210 212 214 216 218
m/z
m/z
mechanism (a→b→c) revealed that upon rapid mixing
mean squared deviation (rmsd) value oꢀ 0.3194 Å ꢀor the
DHP B (F21W) superposition oꢀ protomers A and B is
consistent with highly homologous structures that possess
very little geometric variation over conꢀormational space
(
2 ms) oꢀ a ꢀerric DHP B (F21W) solution [UV–visible
spectrum: 407 (Soret), 505, 535 (sh), 635] with 10 equiv
H O , a new species was observed whose spectral ꢀeatures
2
2
α
[
UV–visible: 419 (Soret), 545, 587 nm] we attributed to
[35–37]. LSQ C superposition oꢀ DHP B (F21W) and
the ꢀerryl-containing intermediate compound ES based
upon our previous assignments oꢀ this reaction intermedi-
ate in WT DHP A [8] and B [9]. The experimental value
WT DHP B yielded a rmsd value oꢀ 0.3491 Å (average
oꢀ subunits), which again shows high structural similarity
globally. Moreover, the hemes align very well despite not
α
oꢀ k ꢀor compound ES ꢀormation varied linearly with
being a variable in the C LSQ alignment (Fig. 5a).
obs
hydrogen peroxide concentration (2.5–25 mol equiv), ꢀrom
which the bimolecular rate constant was determined to be
As opposed to the global comparison, a closer inspec-
tion oꢀ the F21W mutant active site (Fig. 5b) revealed
diꢃerences ꢀrom previous DHP structures (Fig. S2): the
ꢀerric DHP B (F21W) structure shows the absence oꢀ the
usual water molecule as the heme-Fe sixth ligand. Rather,
unique to this DHP structure is the presence oꢀ two water
molecules in the distal pocket that ꢀorm a hydrogen-bond-
ing network with W21, T56, Y38, and the heme propionate
arms. In WT DHP B, T56 resides in a diꢃerent conꢀorma-
4
−1 −1
(
6.68 ± 0.05) × 10 M s , in agreement with the values
4
reported ꢀor WT isoenzymes [DHP A: (3.56 ± 0.02) × 10
−
1
−1
5
−1 −1
M
s
[8]; DHP B: (1.29 ± 0.11) × 10 M
s
[9];
Table 2]. The DHP B (F21W) compound ES intermedi-
ate ꢀurther converted to a new species [UV–visible: 412
−
1
(
Soret), 535, 565 (sh) nm; k = 0.012 ± 0.001 s ]. We
obs
have assigned this as compound RH, the “reversible heme”
state oꢀ dehaloperoxidase that ꢀorms ꢀrom the decay oꢀ com-
pound ES in the absence oꢀ a co-substrate, ꢀor this variant
based upon analogy to the similar, but not identical, species
observed in DHP isoenzymes A [411 (Soret), 530, 564 (sh)
γ
tion, with the O turned away ꢀrom the pocket, interacting
with the E helix. In line with the lack oꢀ a water (sixth)
ligand in DHP B (F21W), the distal histidine, H55, does
not participate in the H-bond network inside the distal
pocket, but instead resides in the solvent-open conꢀorma-
tion interacting with the heme propionate arms and a sul-
ꢀate molecule. The heme was observed in a slightly twisted
and domed orientation, with the Fe lying 0.06 Å below
the heme pyrrole N plane (vs 0.21 Å ꢀor ꢀerric WT DHP
B [18]). Comparing subunit-averaged distances (Table 4)
−
1
nm; k = 0.0167 ± 0.0003 s ] [8] and B [411 (Soret),
obs
−
1
5
54, 599 (sh) nm; k = 0.010 ± 0.001 s ] [9], as well as
obs
in various other DHP mutants [13].
X‑ray crystallographic studies
ε
Non-His tagged DHP B (F21W) was overexpressed, puri-
fed, crystallized, and its structure solved by X-ray diꢃrac-
tion methods. X-ray data collection and refnement statistics
are provided in Table 3. DHP B (F21W) refned to a resolu-
between DHP B (F21W) and WT DHP B, the Fe–H89 N
distance is 0.10 Å longer in the mutant (2.28 vs 2.18 Å)
δ
while the H89 N –L83 carbonyl O distance is shorter by
α
0.14 Å (2.73 vs 2.87 Å). However, the Fe–H89 C dis-
tion oꢀ 1.96 Å, with R
, R
and R values oꢀ 8.4,
tances are comparable (6.15 vs 6.13 Å), suggesting that
the diꢃerences in H89 distances can be ascribed to the side
chain itselꢀ and not the protein backbone. Additionally,
merge
work
ꢀree
1
6.43 and 20.45%, respectively.
DHP B (F21W) crystallized as a homo-dimer in the
α
α
asymmetric unit oꢀ the P2 2 2 space group, consistent
Fe–H55 C distances (8.69 vs 8.67 Å) and Fe–C distances
ꢀor residue 21 (W21:10.65 Å vs F21:10.75 Å) show very
little deviation between mutant and wild-type main chain
distances, resulting in similar heme pocket geometries.
1
1 1
with all previous DHP crystal structures. The protomer
environments were identical, as shown through the global
α
least squared quadratic (LSQ) C superpositions: the root
1
3

JBIC Journal of Biological Inorganic Chemistry
Discussion
A
407
419
1
1
0
0
0
.10
.00
.90
.80
.70
Raw
Fit
Using the previous study oꢀ Dawson and co-workers [24]
as inspiration ꢀor the present work, the F21W mutation
was rationally designed with the intention oꢀ blocking the
TCP
binding site in an eꢃort to increase the selec-
interior
0
5
10
15
20
tively oꢀ DHP as a peroxygenase enzyme at the expense
oꢀ its peroxidase activity. To that end, the DHP B (F21W)
mutant exhibited minimal decreases in peroxygenase
activity oꢀ 1.2–2.1-ꢀold as measured by substrate conver-
sion, whereas the catalytic eꢁciency oꢀ the peroxidase
activity was more than sevenꢀold decreased. Interest-
ingly, Dawson and co-workers showed that the catalytic
eꢁciency ꢀor DHP A (F21W) peroxidase reactivity was
reduced 15-ꢀold compared to WT DHP A [24]. Thus, the
question arose as to why the eꢃect oꢀ the F21W muta-
tion is more muted in DHP B than in DHP A. From a
structural standpoint, the diꢃerences in primary structure
between these two isoenzymes are limited to fve amino
acid substitutions (note—isoenzyme A is listed frst): I/
L9, R/K32, Y/N34, N/S81 and S/G91. These amino acid
substitutions do not perturb the overall structural ꢀold oꢀ
DHP when comparing the two isoenzymes [18], yet they
result in DHP B exhibiting a greater reactivity than DHP
A ꢀor both peroxidase and peroxygenase activities [4, 5,
Time (sec)
5
45
5
87
Wavelength (nm)
B
120
00
4
07
419
1
5
35
2
1
1
0
5
0
5
0
412
5
05 545
80
60
40
20
0
587
4
50
500
550
600
650
700
9
]. However, why the F21W mutation aꢃects DHP A more
3
00 350 400 450 500 550 600 650 700 750 800
than DHP B cannot be addressed by a simple structural
comparison (vide inꢀra).
Wavelength (nm)
1
.8
.6
.4
.2
0
C
To begin to address this question, we perꢀormed
stopped-ꢂow UV/visible spectroscopic studies to assess
whether the ꢀormation oꢀ the compound ES intermediate
in DHP B (F21W) was altered owing to the introduction
oꢀ a redox-active tryptophan residue near the active site.
However, no signifcant diꢃerences in the rates oꢀ ꢀorma-
tion and decay, or in its spectral ꢀeatures, were noted ꢀor
compound ES, suggesting that the introduction oꢀ the
F21W mutation does not alter the electronic structure oꢀ
this reactive intermediate. Similarly, we were concerned
iꢀ the F21W mutation alters the oxidation pathway oꢀ
haloindoles (i.e., causes a switch ꢀrom O atom transꢀer
to electron transꢀer); however, isotope labeling studies
confrmed the peroxygenase mechanism oꢀ DHP remains
intact in DHP B (F21W). Taken together, we conclude
that the F21W mutation does not signifcantly change the
nature oꢀ the catalytic intermediates, and thus does not
alter the mechanisms that have been established in DHP
Ferric (resting)
0
Compound ES
1
Compound RH
0
.8
0
0
0
0.6
0
0
.4
.2
0
0
0.5
20
0
40
60
80
Time (sec)
Fig. 4 Kinetic data obtained by optical spectroscopy ꢀor the reac-
tion oꢀ DHP B (F21W) with H O . a Stopped-ꢂow UV–visible spec-
2
2
tra oꢀ the reaction oꢀ DHP B (F21W) (10 µM) with a ten-ꢀold excess
oꢀ H O at pH 7 (900 scans over 83 s); inset—the single wavelength
2
2
(
407 nm) dependence on time obtained ꢀrom the raw spectra and its
ft with a superposition oꢀ the calculated spectra components; b Cal-
culated spectra oꢀ the three reaction components derived ꢀrom the
SVD analysis: ꢀerric (black), compound ES (red), and compound RH
ꢀ
or peroxidase [2] and peroxygenase [4–6] activities.
Having ruled out changes to the catalytic intermedi-
(
blue); c time dependences oꢀ the relative concentrations ꢀor the three
ates as being responsible ꢀor the diꢃerences in activity
between DHP A (F21W) and DHP B (F21W), we con-
sidered structural (and potentially dynamic) eꢃects oꢀ
the catalytically relevant proximal and distal histidine
components shown in panel B
1
3

JBIC Journal of Biological Inorganic Chemistry
Table 2 Comparison oꢀ UV–visible spectroscopic data ꢀor the ꢀerric, compound ES, and compound RH species in DHP B (F21W) and WT DHP
isoenzymes A and B at pH 7
Enzyme
Ferrica
Compound ES
Compound RH
Reꢀerences
b
DHP B (F21W)
WT DHP A
WT DHP B
407, 505, 535 (sh) 635
407, 504, 538, 635
407, 508, 633
419, 545, 587
420, 545, 585
419, 545, 585
412, 535, 565
411, 530, 564
411, 554, 599
[8]
[9]
a
Values reported are in nm
This work
b
T56
F21
Y38
H55
3
.0
W21
W21
2.9
3.0
Table 3 X-ray data collection and refnement statistics ꢀor DHP B
F21W) (5VLX)
2.8
(
2.9
PDB accession code
5VLX
Heme β edge
Data collection
Wavelength (Å)
Temperature (K)
Space group
Unit-cell parameters (Å)
a
Heme β edge
1.54
100
H89
H89
A
B
P2 2 2
1
1 1
α
Fig. 5 a DHP B (F21W) (silver) C superposition with WT DHP B
(cyan; PDB 3IXF [18]), highlighting the F21/W21 residues. Only one
conꢀormation ꢀor F21 is shown ꢀor clarity. b Distal pocket oꢀ the DHP
B (F21W) mutant, as viewed ꢀrom the heme β edge. There are two
water molecules in the distal pocket, ꢀorming a H-bonding network
with W21, T56, Y38, and the heme propionate arms. All distances
are given in Å
60.48
67.13
67.37
25,631 (1032)
97.8 (80.2)
8.4 (39.5)
25.0 (2.4)
4.5 (3.3)
2.21
b
c
a
Unique reꢂections
Completeness (%)
Rmerge (%)b
I/σ(I)
a
a
a
Table 4 Selected distances (Å) ꢀor DHP B (F21W) and WT DHP B
Redundancy
(
PDB accession 3IXF [18])
3
Vm (Å /Da)
WT DHP Ba
Refnement
DHP B (F21W)
Protomer A Protomer B Protomer A Protomer B
Resolution (Å)
1.80
c
a
a
Rwork (%)
16.62 (21.59)
20.44 (29.75)
2317
ε
Rꢀree (%)d
Fe–H89 N
2.28
2.27
2.74
6.14
8.77
10.63
2.18
2.92
6.14
8.71
10.76
2.17
2.81
6.11
8.63
10.74
δ
H89 N –L83 O 2.71
No. oꢀ protein atoms
No. oꢀ solvent atoms
R.m.s.d ꢀrom ideal geometry
Bond lengths (Å)
α
Fe–H89 C
6.16
8.61
10.67
215
Fe–H55 Cα
Fe–F21/W21
e
0.007
0.926
α
C
Bond angles (°)
η
η
6.50 (Cζ)
6.32 (Cζ)
Fe–F21/W21
Side chain
5.06 (C )
5.03 (C )
Ramachandran plot (%)
Most ꢀavored region
Addl allowed region
ζ2
ζ2
ε1
ε1
5.47 (C ) 5.49 (C ) 6.74 (C ) 6.60 (C )
98.89
1.11
a
For clarity, only distances ꢀor the F21 conꢀormer closest to the heme
are given ꢀor WT DHP B
a
Values in parentheses are ꢀor the highest resolution shell
b
R
= ∑ ∑ [|I (h) − 〈I(h)〉|/∑ ∑ I(h)] × 100%, where I (h) is the
merge
h i i h i i
ith measurement and 〈I(h)〉 is the weighted mean oꢀ all measurements
residues. Our rationale ꢀor such a consideration is that it
has been previously shown that tuning oꢀ DHP activity
may be accomplished by altering (i) the proximal cavity
charge relay [11, 38] that includes histidine (H89), whose
role is to ꢀacilitate dioxygen binding to the ꢀerrous heme
or to support the Fe(IV) oxidation state in compounds
I, ES and II, and (ii) the conꢀormational ꢂexibility and/
oꢀ I(h)
c
R
= ∑|F − F |/∑F × 100%, where FO and FC are the
O C O
work
observed and calculated structure ꢀactors, respectively
d
R
is the R ꢀactor ꢀor the subset (7.8%) oꢀ reꢂections selected
ꢀ
ree
beꢀore and not included in the refnement
e
Root-mean-square deviation
Ramachandran plot created via MolProbity
ꢀ
1
3

JBIC Journal of Biological Inorganic Chemistry
or position oꢀ the catalytically important distal histidine
On the whole, the introduction oꢀ the F21W mutation
causes very little structural perturbation relative to the
(
(
H55) [39–41], which is critical ꢀor enzyme activation
i.e., cleavage oꢀ compound 0 to the activated ꢀerryl spe-
α
wild-type structure. As shown by the C superposition, the
cies). In the context oꢀ the structure–ꢀunction relationship
ascribed to these two residues, we propose the ꢀollow-
ing fve ꢀactors that may individually, or in combination,
account ꢀor the diꢃerential eꢃect oꢀ the F21W mutation in
DHP A and B when compared to the WT isoenzymes: (1)
DHP possesses a Leu-His-Fe charge relay (as opposed to
the typical peroxidase Asp-His-Fe one) whose ꢀunction is
to stabilize the ꢀerryl intermediates through polarization
oꢀ the proximal histidine [2, 11]. In DHP B (F21W), the
heme retains the same position and orientation, which was
unexpected given the increased size oꢀ the tryptophan side
chain compared to that oꢀ phenylalanine. The most notable
diꢃerence between the F21W mutant and WT structure was
observed in the side chain distances ꢀor the mutated residue.
Remarkably, however, the protein backbone and remaining
parts oꢀ the structure were virtually identical in spite oꢀ
the steric diꢃerences imposed by the larger indole ring. As
expected ꢀor the F21W mutation, the tryptophan side chain
is in closer proximity to the heme coꢀactor. Averaging subu-
nits, in WT DHP B the closest atom oꢀ the phenyl ring to the
δ
H89 N –83 carbonyl O distance oꢀ 2.72 Å is shorter by
0
.14 Å than that ꢀound in WT DHP B (2.87 Å), and would
ζ
ε1
better support a ꢀerryl intermediate. (2) The distal histi-
dine is unusually ꢂexible in DHP, being normally ꢀound
in equilibrium between the “open” (5cHS) and “closed”
heme Fe is C at a distance oꢀ 6.41 Å, with C as the next
η
closest atom at 6.67 Å. In the F21W mutant, the C atom
resides 5.05 Å ꢀrom the heme Fe with the next closest atom,
ζ2
(
6cHS, water-bound) states in the WT isoenzymes. It has
C
located 5.48 Å ꢀrom the Fe. Thus, when compared to
been hypothesized [1, 25, 26, 42–44] that altering this
conꢀormational ꢂexibility may aꢃect substrate/co-substrate
binding by gating the accessibility oꢀ the active site pocket
F21 in WT DHP B, W21 has a contact ~ 1.3 Å closer to the
heme Fe. As a consequence, W21 eꢃectively resides in the
TCP
binding site [24] (Fig. 6a), which can be directly
interior
ꢀor entry/exit oꢀ substrate/product molecules or co-sub-
related to the attenuation oꢀ peroxidase reactivity. However,
strate H O [2, 40]. In DHP B (F21W), the distal histidine
peroxidase activity is still observed, which may be explained
2
2
was only ꢀound in the ‘open’ (5cHS) solvent-accessible
conꢀormation, stabilized by H-bonding interactions with
the heme propionate arms and an exogenous sulꢀate ion.
This ‘open’ conꢀormation oꢀ H55 enables the presence oꢀ
active site waters by excluding the H55 ꢀrom the distal
pocket. (3) Although no water ligand was ꢀound to bind
the heme-Fe in both protomers (a ꢀeature that has com-
monly been observed in ꢀerric DHP structures [18, 44]),
two water molecules were ꢀound in the distal pocket. The
introduction oꢀ a nitrogen-containing indole ring ꢀrom the
tryptophan side chain likely increases the polarity oꢀ the
distal pocket, creating a more attractive environment ꢀor
accommodating water molecules in the active site. The
presence oꢀ these distal waters also involves a H-bonding
by the presence oꢀ a second TCP binding site (TCP
).
exterior
This shows that while the TCP
binding site has a sig-
interior
nifcant role in TCP oxidation, it is not solely responsible ꢀor
TCP reactivity, and the TCP binding site may play a
exterior
ꢀunctional role as well through long-range electron transꢀer.
As has been established ꢀor DHP, the presence oꢀ peroxi-
dase substrates within the heme pocket does not ꢀollow the
normally observed peroxidase substrate binding motiꢀ at the
heme γ or δ edge [2]. DHP thus presents an unconventional
example oꢀ observed internal binding oꢀ substrates that
undergo oxidation through a mechanism that is generally
characterized by substrate binding external oꢀ the heme cav-
ity [45]. Internal peroxidase substrate binding is an obvious
consequence oꢀ the unusually large distal cavity presented
in DHP (even though the tertiary structure possesses a
γ
network that includes the O oꢀ the T56 side chain (now
unable ꢀor stabilization oꢀ Helix E), Y38 (site oꢀ radical
ꢀ
ormation in DHP B), a heme propionate arm, and W21.
4) In studies oꢀ the T56 mutants oꢀ DHP [39], a greater
(
W21
W21
conꢀormational ꢂexibility oꢀ the distal histidine was noted.
γ
In DHP B (F21W), it was noted that the O oꢀ the T56 side
TCPext
chain was positioned into the distal pocket instead oꢀ its
normal role oꢀ stabilization oꢀ helix E, which could lead
to altered H55 ꢂexibility. (5) Although no crystal structure
oꢀ DHP A (F21W) is available, calculations showed that
the orientation oꢀ the W21 side chain was directed away
TCPint
4
NP
Heme β edge
Heme β edge
H89
ꢀ
rom the heme iron and toward the δ heme edge [24]. In
A
H89
B
the DHP B (F21W) crystal structure, however, W21 was
observed positioned directly toward the heme-Fe, which
would likely alter the reactivity in comparison to that oꢀ
isoenzyme A mutant.
Fig. 6 DHP B (F21W) mutant distal pocket superposition with TCP
TCP (PDB 4KMV) and TCP (PDB 4KMW) [24]] and
[
interior
exterior
4-NP (PDB 5CHQ [4]) substrates
1
3

JBIC Journal of Biological Inorganic Chemistry
traditional globin ꢀold), which is spacious enough to accom-
modate internal binding oꢀ phenolic peroxidase substrates.
Thus, the peroxidase mechanism oꢀ DHP may in ꢀact exclude
the generally observed requirement oꢀ external substrate
binding, with the two sequential one-electron oxidation steps
can accommodate large substrates (i.e., haloindoles) and
increases in side chain volume (i.e., Phe → Trp) while main-
taining a nearly identical backbone structure. The dominant
change in the side chains appears restricted to the position
oꢀ the distal histidine in that it is ꢀorced into the solvent-
exposed conꢀormation. This change does have ꢀunctional
consequences, although it does not itselꢀ constitute an oꢃ
switch—surprisingly, the activation oꢀ bound H O is robust
[
15] possibly occurring within the distal cavity.
Peroxygenase reactivity, on the other hand, is charac-
terized by atom transꢀer, and to ꢀacilitate this transꢀer the
substrate binding sites have greater constraints on their pos-
sible orientations and distances ꢀrom the activated ꢀerryl
intermediate. W21 only partially overlaps with the 4-NP
binding site [4] (Fig. 6b), specifcally with the nitro group
oꢀ 4-NP, but not with the bulk oꢀ the 4-NP substrate. The
limited steric clash with the nitro group oꢀ 4-NP likely still
permits it to bind in an orientation that is in close proxim-
ity to the activated heme Fe, which in turn would appear to
permit oxygen atom transꢀer as required ꢀor the peroxyge-
nase mechanism. Crystallographic substrate binding sites
oꢀ 5-Br-indole and 7-Br-indole have yet to be obtained,
however, computationally [5] derived binding sites are in
good agreement with spectroscopic data: (i) 5-Br-indole
has been computed to bind in the distal pocket above the
heme Fe, which correlates to the substrate’s perturbation
oꢀ the 5cHS/6cHS heme equilibrium to ꢀavor the 5cHS
species; (ii) the 7-Br-indole computed binding site resides
deeper in the distal pocket, and consequently the binding oꢀ
2
2
and appears to be possible even in a structure where H55 is
ꢀorced into the external position through the combination
oꢀ binding a large substrate and a large side chain. Thus, the
F21W mutant confrms that H O -binding and subsequent
2
2
enzyme activation may still be possible even when H55 is
ꢀorced into the external position. Our studies here thereꢀore
suggest that it may be possible to ꢀurther tune DHP activity
through simultaneous mutagenesis oꢀ multiple active site
residues without the requirement that the distal histidine
be positioned in the internal conꢀormation to serve as the
general acid/base catalyst oꢀ the Poulos–Kraut mechanism.
Acknowledgements This project was supported by NSF CAREER
Award CHE-1150709 and NSF CHE-1609446. Mass spectra were
obtained at the Mass Spectrometry Facility ꢀor Biotechnology at North
Carolina State University. Partial ꢀunding ꢀor the ꢀacility was obtained
ꢀrom the North Carolina Biotechnology Center and the National Sci-
ence Foundation.
7
-Br-indole has little or no eꢃect on the 5cHS/6cHS heme
equilibrium, indicating this substrate binds in an orienta-
tion that does not inhibit water ligation to the heme Fe. In
DHP B (F21W), the additional steric bulk oꢀ the W21 side
chain is in a position to hinder 5-Br-indole binding while
not ꢀully blocking the binding site. This was reꢂected in
the dissociation constant ꢀor 5-Br-indole in DHP B (F21W)
References
1
2
3
4
.
.
.
.
Lebioda L, LaCount MW, Zhang E, Chen YP, Han K, Whitton
MM, Lincoln DE, Woodin SA (1999) Nature 401:445
Franzen S, Ghiladi RA, Lebioda L, Dawson J (2016) Heme peroxi-
dases. The Royal Society oꢀ Chemistry, Cambridge, pp 218–244
Chen YP, Woodin SA, Lincoln DE, Lovell CR (1996) J Biol Chem
(
K = 2960 ± 940 μM; Fig. S1) being 20-ꢀold higher than in
d
2
71:4609–4612
WT DHP B (K = 150 ± 10 μM), which likely results ꢀrom
d
McCombs NL, D’Antonio J, Barrios DA, Carey LM, Ghiladi RA
either weaker binding in the same site as the WT enzyme, or
possibly that binding occurs at a diꢃerent site. In agreement
with the computed binding oꢀ 7-Br-indole showing that this
substrate resides deeper in the distal pocket, the additional
steric bulk oꢀ W21 appears to have a very small eꢃect on
(
2016) Biochemistry 55:2465–2478
5. Barrios DA, D’Antonio J, McCombs NL, Zhao J, Franzen S,
Schmidt AC, Sombers LA, Ghiladi RA (2014) J Am Chem Soc
1
36:7914–7925
6
. McCombs NL, Smirnova T, Ghiladi RA (2017) Catal Sci Technol
7
:3104–3118
7
-Br-indole reactivity, with a decrease oꢀ only 1.2-ꢀold.
7. Osborne RL, Taylor LO, Han KP, Ely B, Dawson JH (2004) Bio-
chem Biophys Res Commun 324:1194–1198
8
.
Feducia J, Dumarieh R, Gilvey LB, Smirnova T, Franzen S,
Ghiladi RA (2009) Biochemistry 48:995–1005
Conclusion
9
.
D’Antonio J, D’Antonio EL, Thompson MK, Bowden EF, Franzen
S, Smirnova T, Ghiladi RA (2010) Biochemistry 49:6600–6616
1
1
0. Thompson MK, Franzen S, Ghiladi RA, Reeder BJ, Svistunenko
DA (2010) J Am Chem Soc 132:17501–17510
The DHP B (F21W) mutant was rationally designed in an
eꢃort to decrease peroxidase reactivity through perturba-
tion oꢀ the peroxidase substrate binding site, with a minimal
1. D’Antonio EL, D’Antonio J, de Serrano V, Gracz H, Thompson
MK, Ghiladi RA, Bowden EF, Franzen S (2011) Biochemistry
50:9664–9680
(
albeit non-zero) consequence on peroxygenase activity. As
1
1
2. D’Antonio J, Ghiladi RA (2011) Biochemistry 50:5999–6011
3. Dumarieh R, D’Antonio J, Deliz-Liang A, Smirnova T, Svis-
tunenko DA, Ghiladi RA (2013) J Biol Chem 288:33470–33482
4. Osborne RL, Sumithran S, Coggins MK, Chen YP, Lincoln DE,
Dawson JH (2006) J Inorg Biochem 100:1100–1108
such, the F21W mutation enables the selective tuning oꢀ
activity in this multiꢀunctional hemoprotein. The results
provide ꢀurther evidence that confrm the plasticity oꢀ the
DHP heme distal cavity: namely, the distal cavity oꢀ DHP
1
1
3

JBIC Journal of Biological Inorganic Chemistry
1
1
5. Osborne RL, Coggins MK, Raner GM, Walla M, Dawson JH
AJ, Moriarty NW, Oeꢃner R, Read RJ, Richardson DC, Richard-
son JS, Terwilliger TC, Zwart PH (2010) Acta Cryst D66:213–221
32. Emsley P, Lohkamp B, Scott WG, Cowtan K (2010) Acta Cryst
D66:486–501
(
2009) Biochemistry 48:4231–4238
6. Davydov R, Osborne RL, Shanmugam M, Du J, Dawson JH, Hoꢃ-
man BM (2010) J Am Chem Soc 132:14995–15004
7. Du J, Sono M, Dawson JH (2010) Biochemistry 49:6064–6069
8. de Serrano V, D’Antonio J, Franzen S, Ghiladi RA (2010) Acta
Cryst D66:529–538
1
1
33. Aꢀonine PV, Grosse-Kunstleve RW, Echols N, Headd JJ, Moriarty
NW, Mustyakimov M, Terwilliger TC, Urzhumtsev A, Zwart PH,
Adams PD (2012) Acta Cryst D68:352–367
1
2
9. Poulos TL, Kraut J (1980) J Biol Chem 255:8199–8205
0. Dunꢀord HB (2016) Heme peroxidases. The Royal Society oꢀ
Chemistry, Cambridge, pp 99–112
34. Chenprakhon P, Sucharitakul J, Panijpan B, Chaiyen P (2010) J
Chem Educ 87:829–831
35. Carugo O, Pongor S (2001) Protein Sci 10:1470–1473
36. Cohen FE, Sternberg MJ (1980) J Mol Biol 138:321–333
37. Chothia C, Lesk AM (1986) EMBO J 5:823–826
38. Sun S, Sono M, Wang C, Du J, Lebioda L, Dawson JH (2014)
Arch Biochem Biophys 545:108–115
2
1. Ortiz de Montellano PR (ed) (2015) Cytochrome p450: structure,
mechanism, and biochemistry. Springer International Publishing,
Switzerland
2
2
2
2
2
2
2. Franzen S, Thompson MK, Ghiladi RA (2012) Biochim Biophys
Acta 1824:578–588
39. Jiang S, Wright I, Swartz P, Franzen S (2013) Biochimica et Bio-
physica Acta (BBA) Proteins and Proteomics 1834:2020–2029
40. Zhao J, de Serrano V, Dumarieh R, Thompson M, Ghiladi RA,
Franzen S (2012) J Phys Chem B 116:12065–12077
41. Carey LM, Gavenko R, Svistunenko DA, Ghiladi RA (2017) Bio-
chimica et Biophysica Acta (BBA) Proteins and Proteomics (in
press)
3. McCombs NL, Moreno-Chicano T, Carey LM, Franzen S, Hough
MA, Ghiladi RA (2017) Biochemistry 56:2294–2303
4. Wang C, Lovelace LL, Sun S, Dawson JH, Lebioda L (2013) Bio-
chemistry 52:6203–6210
5. LaCount MW, Zhang E, Chen YP, Han K, Whitton MM, Lincoln
DE, Woodin SA, Lebioda L (2000) J Biol Chem 275:18712–18716
6. Thompson MK, Davis MF, de Serrano V, Nicoletti FP, Howes BD,
Smulevich G, Franzen S (2010) Biophys J 99:1586–1595
7. Zhao J, de Serrano V, Le P, Franzen S (2013) Biochemistry
42. de Serrano VS, Davis MF, Gaꢃ JF, Zhang Q, Chen Z, D’Antonio
EL, Bowden EF, Rose R, Franzen S (2010) Acta Cryst
D66:770–782
5
2:2427–2439
43. Chen Z, de Serrano V, Betts L, Franzen S (2009) Acta Cryst
D65:34–40
2
2
8. Beers RF Jr, Sizer IW (1952) J Biol Chem 195:133–140
9. Otwinowski Z, Minor W (1997) In: Carter CW Jr, Sweet RM (eds)
Methods in enzymology. Academic, New York, pp 307–326
0. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni
LC, Read RJ (2007) J Appl Crystallogr 40:658–674
44. de Serrano V, Chen Z, Davis MF, Franzen S (2007) Acta Cryst
D63:1094–1101
3
3
45. Kwon H, Moody PCE, Raven EL (2016) Heme peroxidases. The
Royal Society oꢀ Chemistry, Cambridge, pp 47–60
1. Adams PD, Aꢀonine PV, Bunkoczi G, Chen VB, Davis IW, Echols
N, Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW, McCoy
Aꢀliations
Leiah M. Carey · Kyung Beom Kim1,2 · Nikolette L. McCombs · Paul Swartz · Cheal Kim · Reza A. Ghiladi
1
1
3
2
1
2
*
Reza A. Ghiladi
Department oꢀ Fine Chemistry, Seoul National University
Reza_Ghiladi@ncsu.edu
oꢀ Science and Technology, Seoul 139-743, Korea
3
Department oꢀ Structural and Molecular Biochemistry, North
Carolina State University, Raleigh, NC 27695-7622, USA
1
Department oꢀ Chemistry, North Carolina State University,
Raleigh, NC 27695-8204, USA
1
3