Shifting redox states of the iron center partitions CDO between crosslink formation or cysteine oxidation

Article abstract of DOI:10.1016/j.abb.2014.06.001

Cysteine dioxygenase (CDO) is a mononuclear iron-dependent enzyme that catalyzes the oxidation of l-cysteine to l-cysteine sulfinic acid. The mammalian CDO enzymes contain a thioether crosslink between Cys93 and Tyr157, and purified recombinant CDO exists as a mixture of the crosslinked and non crosslinked isoforms. The current study presents a method of expressing homogenously non crosslinked CDO using a cell permeative metal chelator in order to provide a comprehensive investigation of the non crosslinked and crosslinked isoforms. Electron paramagnetic resonance analysis of purified non crosslinked CDO revealed that the iron was in the EPR silent Fe(II) form. Activity of non crosslinked CDO monitoring dioxygen utilization showed a distinct lag phase, which correlated with crosslink formation. Generation of homogenously crosslinked CDO resulted in an ～5-fold higher k_cat/K_m value compared to the enzyme with a heterogenous mixture of crosslinked and non crosslinked CDO isoforms. EPR analysis of homogenously crosslinked CDO revealed that this isoform exists in the Fe(III) form. These studies present a new perspective on the redox properties of the active site iron and demonstrate that a redox switch commits CDO towards either formation of the Cys93-Tyr157 crosslink or oxidation of the cysteine substrate.

Full text of DOI:10.1016/j.abb.2014.06.001

Archives of Biochemistry and Biophysics 558 (2014) 61–69
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Shifting redox states of the iron center partitions CDO between crosslink
formation or cysteine oxidation
⇑
Catherine W. Njeri, Holly R. Ellis
The Department of Chemistry and Biochemistry, 179 Chemistry Building, Auburn University, Auburn, AL 36849, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Cysteine dioxygenase (CDO) is a mononuclear iron-dependent enzyme that catalyzes the oxidation of
Received 18 April 2014
and in revised form 29 May 2014
Available online 11 June 2014
L-cysteine to L-cysteine sulfinic acid. The mammalian CDO enzymes contain a thioether crosslink between
Cys93 and Tyr157, and purified recombinant CDO exists as a mixture of the crosslinked and non cross-
linked isoforms. The current study presents a method of expressing homogenously non crosslinked
CDO using a cell permeative metal chelator in order to provide a comprehensive investigation of the
non crosslinked and crosslinked isoforms. Electron paramagnetic resonance analysis of purified non
crosslinked CDO revealed that the iron was in the EPR silent Fe(II) form. Activity of non crosslinked
CDO monitoring dioxygen utilization showed a distinct lag phase, which correlated with crosslink forma-
Keywords:
Cysteine dioxygenase
Cys–Tyr thioether crosslink
Metalloprotein
tion. Generation of homogenously crosslinked CDO resulted in an ꢀ5-fold higher kcat/K
m
value compared
L
-Cysteine
to the enzyme with a heterogenous mixture of crosslinked and non crosslinked CDO isoforms. EPR anal-
ysis of homogenously crosslinked CDO revealed that this isoform exists in the Fe(III) form. These studies
present a new perspective on the redox properties of the active site iron and demonstrate that a redox
switch commits CDO towards either formation of the Cys93–Tyr157 crosslink or oxidation of the cysteine
substrate.
L
-Cysteine sulfinic acid
Cupin superfamily
Ó 2014 Elsevier Inc. All rights reserved.
Introduction
conserved Glu residue in motif 1 has been replaced with a cysteine
leaving only three His residues to coordinate the active site iron. In
Enzymatic oxidation of
L
-cysteine to
L
-cysteine sulfinic acid
addition to the unique metal coordination environment, CDO also
possesses a rare crosslink involving Cys93 and Tyr157 (residues
numbered according to mammalian CDO enzymes) [8–11].
(
CSA)1 is catalyzed by the mononuclear iron dependent enzyme,
cysteine dioxygenase (CDO) [1–4]. Both atoms of dioxygen are incor-
porated in the thiol group of cysteine to generate the cysteine
sulfinic acid product (Scheme 1) [5]. The product of this reaction is
a branchpoint in the biosynthesis of pyruvate and inorganic sulfate
or taurine [6].
Cysteine dioxygenase is a member of the cupin superfamily.
Proteins within the cupin superfamily have a low overall sequence
similarity, but they share a common b-barrel fold known as the
cupin fold. Members of the cupin superfamily also share two par-
Several studies have been conducted to evaluate the role of the
crosslink in the oxidation of the cysteine substrate. Purified recom-
binant CDO from mammalian systems migrates as two bands when
resolved on SDS–PAGE. This anomalous migration pattern has been
observed in several other proteins including galactose oxidase,
glyoxal oxidase, and Mycobacterium tuberculosis sulfite reductase
(NirA) all of which contain a Cys–Tyr crosslink [12–14]. Mutational
studies and mass spectrometric analysis have identified the faster
migrating species as the isoform that contains the Cys–Tyr cross-
link and the slower migrating species as the isoform that lacks
the crosslink [15–16]. The kinetic and spectroscopic properties of
CDO that have been reported so far have involved a heterogenous
mixture of the crosslinked and non crosslinked CDO isoforms. In an
effort to generate the homogenously crosslinked isoform, the het-
erogenous mixture of crosslinked and non crosslinked CDO iso-
forms was incubated with cysteine and ferrous iron under
aerobic conditions [15]. Crosslinked CDO once formed was shown
to possess increased activity several fold over variants of CDO lack-
ing the Cys93–Tyr157 crosslink; however, the crosslink was not
tially conserved cupin motifs, GX
5 6
HXHX3–6EX G (motif 1) and
GX5–7PXGX HX N (motif 2) [7]. The two His and Glu residues from
2
3
motif 1 and His from motif 2 coordinate the active site metal in
most proteins within the cupin superfamily. For CDO, the
⇑
Corresponding author. Fax: +1 334 844 6959.
E-mail address: ellishr@auburn.edu (H.R. Ellis).
Abbreviations used: CDO, cysteine dioxygenase; SDS–PAGE, sodium dodecyl
1
sulfate–polyacrylamide gel electrophoresis; EPR, electron paramagnetic resonance;
ICP-AES, inductively coupled plasma-atomic emission spectroscopy; CSA, cysteine
sulfinic acid; MCD, magnetic circular dichroism.
http://dx.doi.org/10.1016/j.abb.2014.06.001
0
003-9861/Ó 2014 Elsevier Inc. All rights reserved.

6
2
C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
O
S
Expression and purification of C93S, Y157F, and wild-type CDO
SH
O
The cDNA gene for rat CDO was cloned into the pET21a expres-
sion vector, and used to generate C93S CDO as previously described
O2
[
19]. Primers for the Y157F CDO variant were designed as 29 base
CDO
oligonucleotides replacing the wild-type Tyr codon with TTT. The
Tyr to Phe substitution was confirmed by DNA sequence analysis.
The respective plasmid was transformed into Escherichia coli
BL21 (DE3) competent cells for protein expression and stored as
glycerol stocks at À80 °C. Expression and purification of the wild-
type and variants of CDO were carried out as previously described
with minor modifications [19]. The cell pellet was resuspended
with 25 mM HEPES, pH 7.5, and 10% glycerol (v/v) supplemented
CO2
H N
3
CO2
L-cysteine sulfinic acid
H N
3
L
-cysteine
Scheme 1. Cysteine oxidation reaction by CDO.
with 2
precipitation step was supplemented with 20% ammonium sulfate
w/v) and applied onto a Phenyl Sepharose™ 6 Fast Flow column.
Protein was eluted using a linear gradient from 20% to 15% ammo-
nium sulfate in 25 mM HEPES buffer, pH 7.5, and 10% glycerol. Pure
protein fractions were pooled and concentrated by precipitation in
lg/mL lysozyme. The supernatant following the nucleic acid
absolutely required for activity [15]. In a separate study, a 1:3 ratio
of crosslinked to non crosslinked CDO was generated through
anaerobic protein expression and separation by column chroma-
tography [17]. The activity assays showed a slow phase associated
with the generation of the crosslink followed by a faster phase that
correlated with increased activity due to crosslink formation. The
biphasic trace was thought to be associated with an uncharacter-
ized intermediate generated during crosslink formation and
enzyme activation [17]. Results from recent spectroscopic and
computational studies support a role for the crosslink in stabilizing
the substrate-bound enzyme [18]. Attempts to obtain the fully non
crosslinked isoform of CDO have been unsuccessful, and previous
studies to evaluate crosslink formation have been performed with
recombinant enzyme that contains a significant amount of the
crosslinked isoform [15,17]. Therefore, the non crosslinked and
the crosslinked CDO isoforms have been difficult to evaluate inde-
pendently as both species are present following purification of the
recombinant enzyme. The current study presents a protocol for
expressing the homogenously non crosslinked form of CDO using
a cell permeable metal chelator. The ability to obtain homogenous-
ly non crosslinked CDO provides a more thorough investigation of
the role of the crosslink in catalysis. Kinetic and spectroscopic com-
parisons of the non crosslinked and crosslinked isoforms of CDO
demonstrate a distinct mechanistic role for the Cys93–Tyr157
crosslink in the active site of CDO. Herein, we propose a substrate
activation mechanism in which CDO utilizes the ferric iron to acti-
vate cysteine for reaction with dioxygen. These studies provide a
new perspective on the redox state of the active site iron in
Cys–Tyr crosslink formation and the mechanism of cysteine sub-
strate oxidation.
(
7
0% ammonium sulfate. The protein pellet was resuspended in the
desired volume of 25 mM HEPES buffer, pH 7.5, 100 mM NaCl, and
0% glycerol. The iron loading and dialysis steps were carried
out as previously reported [19]. Protein samples were divided
1
into 500 lL aliquots, flash frozen in liquid nitrogen and stored at
À80 °C.
Expression and purification of non crosslinked CDO
E. coli strain BL21 (DE3) containing the wild-type CDO plasmid
was grown in Luria–Bertani broth containing 0.1 mg/mL ampicillin
with vigorous shaking at 37 °C. When the culture reached an A600
of 0.6–0.8, 1,10-phenanthroline (100 mM dissolved in 100 mM
HCl) was added to a final concentration of 100 lM [20]. The culture
was allowed to grow for an additional 15 min after which the
expression of CDO was induced with 0.4 mM IPTG. The culture
was incubated at 22 °C for an additional 6 h.
Cells were harvested by centrifugation at 3350 Â g at 4 °C and
the cell paste was resuspended in 100 mL of 25 mM HEPES, pH
7
.5, and 10% glycerol supplemented with 2
lg/mL lysozyme. The
cell slurry was lysed by sonication followed by centrifugation at
7
830 Â g at 4 °C for 20 min. Streptomycin sulfate (1.5% w/v) was
added to the resulting supernatant to precipitate nucleic acids
and the solution was stirred for 1 h at 4 °C. The precipitated nucleic
acids were removed by centrifugation at 7830 Â g at 4 °C for
2
0 min. Ammonium sulfate (20% w/v) was added to the superna-
tant and the solution was stirred briefly in the cold room prior to
loading onto a Phenyl Sepharose™ Fast Flow column equilibrated
in 25 mM HEPES buffer, pH 7.5, 20% ammonium sulfate, and 10%
glycerol. Following a column wash, the protein was eluted using
a linear gradient of 20–15% ammonium sulfate in 25 mM HEPES,
pH 7.5, and 10% glycerol. The purest fractions as assessed by
SDS–PAGE were pooled and concentrated by precipitating with
70% ammonium sulfate. The pellet was resuspended in the desired
volume of 25 mM HEPES, pH 7.5, 100 mM NaCl, and 10% glycerol.
The resulting solution was dialyzed twice against 2 L of the resus-
pension buffer to remove residual 1,10-phenanthroline, and the
protein concentration determined by UV–visible absorption spec-
troscopy at 280 nm using a molar extinction coefficient of
25,440 M cm [19]. The purified CDO enzyme was iron reconsti-
tuted with a molar ratio of 1.2:1 ferrous ammonium sulfate to pro-
tein, and dialyzed twice against one liter of 25 mM HEPES, pH 7.5,
100 mM NaCl, and 10% glycerol. Aliquots of the protein samples
were flash frozen in liquid nitrogen and stored at À80 °C. For
clarity, the enzyme isoform lacking the crosslink will be referred
to as non crosslinked CDO, and the enzyme isoform with the
C93–Tyr157 crosslink will be referred to as crosslinked CDO.
Materials and methods
Materials
4
-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES),
-cysteine, -ascorbate, 1,10-phenanthroline, ammonium sulfate,
ampicillin, streptomycin sulfate, lysozyme, sodium dithionite,
potassium chloride (KCl), isopropyl-b- -thiogalactoside (IPTG),
L
L
D
and brilliant blue R were purchased from Sigma (St. Louis, MO).
Sodium dodecyl sulfate (SDS) was purchased from Biorad
(
Hercules, CA). Glycerol and sodium chloride were purchased from
Macron Fine Chemicals (Center valley, PA). Pfu Turbo DNA poly-
merase was purchased from Agilent (La Jolla, CA). Oligonucleotide
primers were purchased from Invitrogen (Carlsbad, CA). Phenyl
Sepharose™ 6 Fast Flow (high sub) was purchased from GE Health-
care Biosciences, (Uppsala, Sweden). Amicon Ultrafree-MC Centrif-
ugal Filter Devices (10 kDa MWCO) and Amicon Ultra-4 Centrifugal
Filter Devices (10 kDa MWCO) were purchased from Millipore
À1
À1
(
Billerica, MA). Difco-brand Luria–Bertani (LB) media was pur-
chased from Becton, Dickinson and Company (Sparks, MD).

C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
63
Quantitation of iron
Analysis of the cysteine sulfinic acid product
The total iron content in all enzyme preparations was deter-
mined using inductively-coupled plasma-atomic emission spec-
The rate of dioxygen consumption was correlated with the rate
of cysteine sulfinic acid (CSA) formation using LC–MS. The reaction
conditions for dioxygen utilization were similar to the steady-state
kinetic experiments. Samples were analyzed for the concentration
of dioxygen consumed within one minute, and the reactions were
troscopy (ICP-AES). Samples of CDO (10
HEPES, pH 7.5 at a final volume of 600
l
M) prepared in 25 mM
lL were analyzed using a
Perkin-Elmer Optima 7300 DV ICP-AES (Perkin-Elmer Life Sciences,
Fremont, CA). Commercial iron standards for ICP-AES were used to
generate the standard curve and the concentration of iron in each
of the prepared protein samples relative to the protein concentra-
tion was determined. Distilled and deionized water and 25 mM
HEPES buffer, pH 7.5 were used as controls. The results are an
average of three separate experiments.
terminated by the addition of 1 lL of formic acid (90% v/v). Mass
analysis was performed using an Ultra Performance LC System
(ACQUITY, Waters Corp., Milford, MA, USA) coupled with a quadru-
pole time-of-flight mass spectrometer (Q-Tof Premier, Waters)
+
with electrospray ionization (ESI) in ESI -MS mode. Each sample
Ò
was initially applied to a C18 column (ACQUITY UPLC BEH C18,
1
.7
l
m, 2.1 Â 50 mm, Waters) for separation with a linear gradient
prior to mass analysis. Solution A contained 5% acetonitrile and
Correlating crosslink formation with non crosslinked CDO activity
0
0
.1% formic acid, and solution B contained 95% acetonitrile and
.1% formic acid. After a 10 min wash with 95% solution A, the
The specific activity of non crosslinked CDO enzyme was mea-
sured by the rate of dioxygen utilization with a Clark-type oxygen
electrode (Hansatech, Inc., Norfolk, United Kingdom). The reactions
applied samples were separated with a linear gradient of 95% A
to 5% A over 6 min (flow rate 150 L/min). Ions of interest were
analyzed for mass accuracy and elemental composition and
the respective ions were quantified by computing the intensity of
the chromatograms. The concentration of CSA formed was deter-
mined by comparison to a standard curve constructed with known
l
consisted of 2 lM CDO and 1 mM ascorbate in 25 mM HEPES buf-
fer, pH 7.5. Reactions were initiated by the addition of varying con-
centrations of cysteine (0–100 mM) to the reaction mixture in a
final volume of 1 mL. Reactions monitored the decrease in the con-
centration of dioxygen in the reaction chamber over a period of
quantities of CSA (10 lM to 5 mM). All the values obtained are an
average of three separate experiments.
1
0 min after which one of the substrates became limiting and the
kinetic traces leveled off. Samples used in the kinetic investigations
were immediately ultrafiltered using Amicon Ultrafree-MC centrif-
ugal filter devices (10 kDa MWCO). Protein samples retained in the
filtration devices were further analyzed for crosslink formation by
SDS–PAGE.
EPR Spectroscopy
Electron paramagnetic resonance (EPR) spectra at X-band fre-
quency (9.38 GHz) were recorded on a Bruker EMX spectrometer
(Bruker Biospin Corporation, Billerica, MA). Cooling was performed
with an Oxford Instruments ESR 900 flow cryostat and an ITC4
temperature controller. Samples were prepared for EPR analysis
by diluting respective protein samples to a final concentration of
Generation of homogenously crosslinked form of CDO
8
1
0–100
l
M in 25 mM HEPES buffer, pH 7.5, 100 mM NaCl, and
The crosslinked form of CDO was generated by incubating non
0% glycerol in a final volume of 300
lL. For samples where cys-
crosslinked CDO (5
lM) in the presence of 100 mM cysteine and
teine was included, 10 mM cysteine was added to diluted protein
samples in a final volume of 300 L. For anaerobic studies, protein
2
5 mM HEPES buffer, pH 7.5 in 500
lL total volume. The reactions
l
were incubated at 37 °C for a period of 15 min with moderate shak-
ing after which all the reactions were combined and ultrafiltered
using Amicon Ultra-4 Centrifugal Filter Units (10 kDa MWCO).
The flow through was discarded and the protein samples retained
in the filtration devices were washed three times with 25 mM
HEPES buffer, pH 7.5, and 10% glycerol. Complete formation of
the crosslinked enzyme was confirmed by SDS–PAGE and the
crosslinked protein was assayed for activity and used in EPR
analysis.
samples were first prepared in crimp top vials and sealed with an
aluminum cap with a rubber septum. The protein was deoxygen-
ated with 20 min of alternate vacuuming and equilibration with
high purity argon on an ice slurry. The samples were then trans-
ferred into a glovebox maintained under an atmosphere of 95%
nitrogen gas and 5% hydrogen gas. Deoxygenated buffer was used
to prepare samples for EPR measurements prior to being trans-
ferred to an EPR tube. All spectra were recorded using the follow-
ing instrument settings: microwave frequency, 9.38 GHz;
4
microwave power, 1.99 mW; receiver gain, 2 Â 10 ; modulation
frequency, 100 kHz; modulation amplitude, 6 G; time constant,
Steady-state kinetic analyses of homogenously crosslinked CDO
3
27.68 ms; sweep time, 335.5 s. The EPR spectra were recorded
as the sum of two accumulations and were normalized by subtract-
ing the spectrum of buffer used to prepare the samples. Spin quan-
titation was performed using 10 mM copper perchlorate as the
standard (10 mM CuSO , 2 mM NaClO , 10 mM HCl) and the con-
4 4
centration of iron was calculated relative to the amount of iron
bound as determined by ICP-AES.
Steady-state kinetic investigations of homogenously cross-
linked CDO were determined using an oxygen electrode where
the rate of dioxygen utilization was monitored. The standard assay
contained 1
and in the absence of 1 mM ascorbate in 25 mM HEPES buffer, pH
.5 at 37 °C. Reactions were initiated by the addition of cysteine
10 M to 10 mM). Initial velocities were obtained within the first
0 s of the reaction and plotted against substrate concentration.
lM crosslinked CDO protein incubated in the presence
7
(
l
1
Results
The steady-state kinetic parameters of wild-type, C93S, and
Y157F CDO proteins were determined using 2 M protein in the
l
Preparation of homogenously non crosslinked CDO
presence of 1 mM ascorbate. The data were fit to the Michaelis–
Menten equation using KaleidaGraph™ software to obtain the
steady-state kinetic parameters which were calculated as the aver-
age of three separate experiments. The steady-state kinetic param-
eters were corrected for the amount of iron bound in each enzyme
preparation.
The similar physical properties between the crosslinked and
non crosslinked isoforms of CDO make it difficult to isolate each
isoform for independent study. Prior attempts to obtain the
homogenous CDO isoforms have proven unsuccessful [17]. A pro-
cedure for isolating the homogenously non crosslinked form of

6
4
C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
Steady-state kinetic parameters of crosslinked CDO
kDa
1
2
3
4
6
6.2
Following the observation that formation of the Cys93–Tyr157
crosslink correlated with increased catalytic activity of CDO, the
steady-state kinetic parameters of the homogenously crosslinked
enzyme were evaluated. Crosslinked CDO was generated by incu-
bating samples containing non crosslinked CDO with 100 mM cys-
teine. SDS–PAGE analysis confirmed the formation of the crosslink
in the prepared protein samples. Enzymatic activity of the cross-
linked CDO enzyme was determined by monitoring the rate of
dioxygen utilization in the absence and in the presence of 1 mM
ascorbate at pH 7.5 (Fig. 3). Including an external reductant in
kinetic investigations of CDO containing both isoforms improved
the catalytic activity of the enzyme [19]. However, crosslinked
CDO was active even in the absence of an external reductant
4
5
1
3
2
1
1.5
4.4
(
Table 1). Similar results were obtained with crosslinked CDO
Fig. 1. Analysis of non crosslinked and mixed wild-type CDO by SDS–PAGE. (1)
Molecular weight marker, (2) purified wild-type CDO containing non crosslinked
and crosslinked isoforms, (3) purified non crosslinked CDO reconstituted with iron,
formed from CDO containing both isoforms (data not shown).
Therefore the external reductant was only necessary when both
CDO isoforms were present. The kinetic parameters were corrected
(
4) purified iron free non crosslinked CDO. Acrylamide gels (12%) were run with
samples pretreated by boiling for 3 min in sample buffer containing 2% SDS and 5%
-mercaptoethanol.
m
for the amount of iron bound and the kcat/K value in the absence
2
À1
À1
of ascorbate was 10,882 ± 276 mM min , which is a ꢀ40-fold
increase compared to the kcat/K value previously reported for
m
CDO was developed in order to provide a comprehensive investiga-
tion on the mechanistic role of the crosslink in cysteine oxidation.
This method involved addition of 1,10-phenanthroline to the cell
culture shortly before inducing the cells for protein expression so
that the enzyme was not exposed to free iron during the expres-
sion and purification processes [20]. The purified CDO enzyme
resolved as a single band on SDS–PAGE, which correlated with
the slower migrating isoform in CDO that lacks the Cys93–
Tyr157 crosslink (Fig. 1). The catalytic activity of CDO depends
on the availability of iron bound in the active site. The amount of
iron in non crosslinked CDO samples was <0.1% following purifica-
tion. It was therefore necessary to reconstitute purified non cross-
linked CDO with iron. Ferrous ammonium sulfate was added to the
purified non crosslinked enzyme in a 1.2:1 (Fe:protein) molar ratio.
The iron loaded protein was then dialyzed twice against an iron
free buffer to remove any unbound iron. Results from ICP-AES
experiments revealed that the iron reconstituted purified non
crosslinked CDO isoform contained ꢀ50% iron. The amount of
bound iron in non crosslinked CDO was comparable to the amount
of bound iron in CDO containing both isoforms.
CDO containing both isoforms with the addition of ascorbate to
the reaction [19].
The reaction catalyzed by CDO requires that one equivalent of
dioxygen reacts with cysteine to generate cysteine sulfinic acid
(CSA). Samples analyzed for dioxygen utilization were also ana-
lyzed for CSA production by mass spectrometry to confirm that
the concentration of dioxygen consumed was coupled with CSA
formation. The specific activity of crosslinked CDO enzyme was
À1
À1
5043 ± 242 nmol of dioxygen consumed min mg of enzyme
which corresponded with the production of 4783 ± 798 nmol of
À1
À1
CSA min mg of enzyme (Table 2). The ꢀ1:1 ratio for the cou-
pling of dioxygen utilization and CSA product formation suggest
that the increased activity observed for crosslinked CDO is due to
oxidation of the cysteine substrate. A similar coupling of specific
activities for dioxygen utilization and CSA production was also
observed for CDO containing both isoforms, but the overall values
were lower than those obtained for crosslinked CDO (Table 2).
Contribution of Cys93 and Tyr157 residues to the catalytic activity of
CDO
Activity of non crosslinked CDO
In order to further evaluate the role of the crosslink in cysteine
substrate oxidation, steady-state kinetic investigations were per-
formed with the Cys93 and Tyr157 CDO variants for comparison
with crosslinked CDO and CDO containing both isoforms. Substitu-
tion of Cys93 to serine and Tyr157 to phenylalanine yielded one
isoform of CDO, which correlated with non crosslinked CDO when
analyzed by SDS–PAGE (Suppl. Fig. 2). The C93S CDO variant con-
tained 39 ± 2% iron bound similar to wild-type CDO, while Y157F
CDO contained lower levels of iron bound at 23 ± 1%. The kinetic
parameters for C93S and Y157F CDO were corrected for the
amount of iron bound (Table 3). The C93S CDO enzyme still pos-
Enzymatic activity of the non crosslinked CDO isoform was
investigated using a Clark-type oxygen electrode and the rate of
dioxygen consumption at varying cysteine concentrations (0–
1
00 mM) was monitored. The concentration of dioxygen in the
reaction chamber was recorded as a function of time (Fig. 2A).
Interestingly, there was an initial lag phase prior to the increase
in rate associated with oxygen utilization. Protein samples from
the activity assays analyzed on SDS–PAGE showed an observable
increase in the amount of crosslink formed with increasing cys-
teine concentration (Fig. 2B). At 80–100 mM cysteine, the homog-
enously non crosslinked CDO isoform in the reaction had been
converted to the homogenously crosslinked isoform in less than
five minutes. ICP-AES experiments were performed to investigate
the amount of iron present following formation of the crosslinked
CDO isoform. Homogenously crosslinked CDO contained ꢀ50% iron
suggesting that complete conversion to the crosslinked isoform
occurred with substoichiometric amounts of iron. However, iron
was required for crosslink formation as non crosslinked CDO that
had not been reconstituted with iron was unable to form the
crosslink when incubated with cysteine under aerobic conditions
sessed activity although the kcat/K
CDO containing both isoforms, and 27-fold lower than crosslinked
CDO in the absence of ascorbate. A comparable decrease in the kcat
value was also observed for the Y157F CDO variant with an
value compared to CDO contain-
m
value was ꢀ5-fold lower than
/
K
m
ꢀ3.5-fold decrease in the kcat/K
m
m
ing both isoforms and a 19-fold reduction in the kcat/K value com-
pared to crosslinked CDO.
EPR Analysis of Non Crosslinked and Crosslinked CDO
The coordinated iron in the resting form of CDO containing both
isoforms was previously shown to exist in the high spin ferric state
(Suppl. Fig. 1).

C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
65
A
350
300
250
200
150
100
+Cys
[
Cys]
0
1
mM
mM
10 mM
20 mM
4
8
0 mM
0 mM
50
0
100
200
300
400
500
Time, s
B
Cysteine 0
1
10 20 30 40 50 80 100
(mM)
Non crosslinked CDO
Crosslinked CDO
Fig. 2. Dioxygen utilization of non crosslinked CDO and crosslink formation. (A) The plot represents the decrease in dioxygen concentration with time at different cysteine
concentrations. The representative kinetic traces show dioxygen utilization at 0–80 mM cysteine. Reaction mixtures contained CDO (2 M) and ascorbate (1 mM) in 25 mM
l
HEPES, pH 7.5 at 37 °C. Assays were initiated by the addition of a range of cysteine concentrations (0–80 mM). (B) Analysis of crosslink formation during catalysis by SDS–
PAGE. Protein samples used in the activity analysis of the non crosslinked CDO enzyme were ultrafiltered with Amicon Ultrafree-MC centrifugal filter devices (10 kDa MWCO)
Acrylamide gels (12%) were run with samples pretreated by boiling for 3 min in sample buffer containing 2% SDS and 5% 2-mercaptoethanol.
Table 1
3000
2000
1000
0
Steady-state kinetic parameters of crosslinked and wild-type CDO.a
b
Crosslinked CDO
+) Ascorbate
Wild-type CDO
(
(À) Ascorbate
(+) Ascorbate
À1
k
cat (min
Km (mM)
(mM min
)
1697 ± 39
0.22 ± 0.02
7714 ± 656
2449 ± 4
0.23 ± 0.01
10,882 ± 276
128 ± 4
0.06 ± 0.01
1995 ± 324
À1
À1
kcat/K
m
)
a
The reported kinetic parameters were adjusted for iron content.
Wild-type CDO containing non crosslinked and crosslinked isoforms.
b
Table 2
Specific activity of dioxygen utilization and cysteine sulfinic acid production of
crosslinked and wild-type CDO.
Dioxygen utilization
(nmol/min/mg)
Cysteine sulfinic acid
production (nmol/min/mg)
0
2
4
6
8
10
12
Crosslinked CDO
Wild-type CDO
5043 ± 242
1007 ± 50
4783 ± 798
1200 ± 15
[L-cysteine], mM
a
a
Fig. 3. Steady-state kinetic traces of crosslinked CDO enzyme. The steady-state
Wild-type CDO containing non crosslinked and crosslinked isoforms.
kinetic traces were obtained by measuring the utilization of dioxygen with a Clark-
type oxygen electrode. Reaction mixtures contained CDO (1
pH 7.5 at 37 °C in the absence (d) and presence (s) of ascorbate (1 mM). Assays
were initiated by the addition of a range of cysteine concentrations (10 M to
0 mM). Each point is the average of at least three separate experiments.
lM) in 25 mM HEPES,
l
as there was no observable EPR signal (Fig. 4, black dashed trace).
Addition of cysteine under aerobic conditions resulted in an
increase in intensity of the EPR signal at g = 4.3 arising from a high
spin ferric iron (S = 5/2) (Fig. 4, black solid trace). Analysis of cross-
linked CDO revealed a sharp EPR signal at g = 4.3 arising from an
S = 5/2 spin system (Fig. 5, black trace). This signal was assigned
to high spin ferric iron bound in the active site of crosslinked
CDO, which accounted for more than 90% of the total iron deter-
mined by double integration of the EPR signal. Contrary to what
1
with an EPR signal observed at g = 4.3 [19]. The addition of cysteine
to oxidized CDO leads to an increase in the signal intensity, and
was attributed to the formation of cysteine-bound Fe(III)-CDO
[
21]. Interestingly, the redox state of the iron in non crosslinked
CDO was found to exist primarily in the EPR-silent ferrous form

6
6
C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
Table 3
Steady-state kinetic parameters and percent iron content of wild-type, C93S, and Y157F CDO.a
À1
À1
À1
kcat (min
)
K
m
(mM)
k
cat/K
m
(mM min
)
% Iron bound
Wild-type CDOb
C93S CDO
Y157F CDO
128 ± 4
45.9 ± 2.4
28.7 ± 1.5
0.06 ± 0.01
0.12 ± 0.02
0.053 ± 0.005
1995 ± 324
399 ± 89
574 ± 64
39 ± 1%
39 ± 2%
23 ± 1%
a
The reported kinetic parameters were adjusted for iron content.
Wild-type CDO containing non crosslinked and crosslinked isoforms.
b
1000
1200
1400
1600
1800
2000
1000
1200
1400
1600
1800
2000
Gauss
Gauss
Fig. 4. X-band EPR spectra of non crosslinked CDO. Non crosslinked CDO (- - -); non
crosslinked CDO with 10 mM cysteine added under aerobic conditions (—). Spectra
Fig. 5. X-band EPR spectra of crosslinked CDO samples. Crosslinked CDO (black
trace); crosslinked CDO with 10 mM cysteine added under aerobic conditions (red
trace); crosslinked CDO with 10 mM cysteine added under anaerobic conditions
were taken with 100
lM protein in 25 mM HEPES buffer, pH 7.5, 100 mM NaCl, and
1
0% glycerol. All spectra were recorded at 9.38 GHz, with a field modulation
(blue trace). Spectra were taken with ꢀ80 lM protein in 25 mM HEPES buffer, pH
frequency of 100 kHz and modulation amplitude of 6 G. All spectra were recorded
at 8–10 K.
7.5, 100 mM NaCl, and 10% glycerol. All spectra were recorded at 9.38 GHz, with a
field modulation frequency of 100 kHz and modulation amplitude of 6 G. All spectra
were recorded at 8–10 K.
is observed for CDO containing both isoforms, addition of cysteine
under aerobic conditions did not lead to significant changes in the
signal intensity (Fig. 5, red trace). The change in the observed spec-
trum between the non crosslinked and crosslinked isoforms sug-
gests that the ferrous iron in the non crosslinked is oxidized to
the ferric form through formation of the crosslink. The EPR signal
intensity at g = 4.3 for crosslinked CDO in the presence of cysteine
was also observed under anaerobic conditions (Fig. 5, blue trace).
It was previously reported that formation of the crosslink in CDO
requires the presence of ferrous iron, L-cysteine, and dioxygen
[15–17]. Because iron is required for crosslink formation, the Fe(II)
chelator 1,10-phenanthroline was added to the cell culture shortly
before induction. SDS–PAGE analysis of the purified protein
revealed the presence of a single isoform of CDO correlating with
protein lacking the crosslink. This is the first time that CDO devoid
of the Cys93–Tyr157 crosslink has been isolated and purified to
homogeneity. Reconstitution of the apo non crosslinked enzyme
with ferrous iron yielded a protein with ꢀ50% iron, which is com-
parable to the value obtained for CDO containing both isoforms.
The amount of iron bound in the active site of CDO ranges from
10% to 68%, which suggests that the 3-His binding site does not
bind iron as tightly as the 3-His/1 carboxylate binding site present
in some cupin metalloproteins [11,22].
There was no EPR signal observed for purified non crosslinked
CDO, which suggested that non crosslinked CDO stabilizes the fer-
rous redox state. With the addition of cysteine, the EPR spectrum
gave a sharp signal at g = 4.3 indicative of high spin ferric iron.
The iron center of CDO containing both isoforms was previously
shown to exist in the high-spin ferric state, and a sharp EPR signal
was observed with the addition of cysteine [19,21]. The increase in
the signal intensity was associated with cysteine coordinating to
the metal center [21]. The change in the EPR signal with the
addition of cysteine to non crosslinked CDO may suggest direct
Discussion
Prior to this study, probes to evaluate the role of the Cys93–
Tyr157 crosslink in CDO have involved the comparative analysis
of variants of CDO unable to form the crosslink and CDO containing
a heterogenous mixture of both the crosslinked and non cross-
linked CDO isoforms. Experiments to separate the crosslinked
and non crosslinked isoforms of CDO have been marginally suc-
cessful [17]. Consistent with previous results, substitution of
Cys93 to serine and Tyr157 to phenylalanine diminished but did
not abolish the catalytic activity of CDO [11,15]. Evaluating the role
of the crosslink through substitution of the amino acids involved in
crosslink formation could be problematic since these residues are
located in the active site and likely play an integral role in catalytic
events. To fully characterize the crosslinked and non crosslinked
isoforms, the two isoforms needed to be evaluated independently.

C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
67
coordination of the substrate to the active site iron or a structural
rearrangement of the active site either due to substrate binding or
the presence of the crosslink. Taken together, the non crosslinked
CDO enzyme stabilizes the ferrous redox state, but the iron is oxi-
dized to Fe(III) upon exposure to cysteine and dioxygen to generate
the crosslink. The requirement that reduced iron is needed for
crosslink formation, may explain why CDO containing both iso-
forms purified in the ferric redox state requires ascorbate in the
reaction [19].
Oxygen electrode measurements revealed an apparent lag
phase prior to the observed decrease in dioxygen concentration
over time. Analysis of the protein samples from the activity mea-
surements by SDS–PAGE showed that increasing the concentration
of cysteine in the assay correlated with increased crosslink forma-
tion. Therefore, the initial lag phase represents the generation of
the Cys–Tyr crosslink prior to cysteine oxidation. Once the cross-
link is generated the rate of dioxygen utilization is stoichiometric
with cysteine sulfinic acid production, suggesting that the
observed rate is correlated with cysteine substrate oxidation. An
apparent biphasic response to oxidation of the cysteine substrate
was recently reported in which an increase in CDO crosslink forma-
tion correlated with an increase in the rate of cysteine sulfinic acid
formation after one hour incubation with cysteine in the presence
of dioxygen [17]. However, formation of the Cys93–Tyr157 cross-
link in CDO is complete in less than 10 min and therefore does
not require lengthy incubation times.
Steady-state kinetic evaluation of crosslinked CDO gave compa-
rable results in the presence and absence of ascorbate. The active
redox state of CDO has been reported to be Fe(II), and CDO contain-
ing both isoforms was shown to have enhanced activity with the
addition of a reductant [19]. Surprisingly, results from EPR analysis
of crosslinked CDO showed that the active site iron was entirely in
the ferric redox state. Furthermore, there was no observable
change in signal intensity when cysteine was added under aerobic
or anaerobic conditions. This data suggests that the active site
geometry of crosslinked CDO may not be altered upon addition
of cysteine either under anaerobic or aerobic conditions. This fixed
intermediate and cysteine thiolate (Scheme 2, III). Nucleophilic
attack of the cysteine thiolate on the tyrosyl radical intermediate
would generate a crosslinked radical intermediate (Scheme 2, IV).
Loss of a proton and electron transfer from the crosslinked radical
intermediate reduces Fe(III) to Fe(II) forming crosslinked CDO
(Scheme 2, V). Protonation of the Fe(II)-hydroperoxy intermediate
would lead to the release of hydrogen peroxide, and water coordi-
nation to the iron would release the Cys substrate (Scheme 2, V).
The reduced iron is subsequently oxidized to Fe (III) in the pres-
ence of molecular oxygen to form the oxidized crosslinked enzyme
observed in EPR studies (Scheme 2, VI). The amount of crosslink
formed is dependent on the cysteine substrate concentration.
Although the described mechanism does not account for this
observed dependence, crosslinked Tyr157 has been proposed to
properly position the cysteine substrate in the active site. There-
fore the cysteine substrate affinity is likely decreased in non cross-
linked CDO because Tyr157 is not in the correct orientation to
facilitate substrate binding.
It is evident from the steady-state kinetic studies that formation
of the crosslink and oxidation of the bound iron to the ferric form
are two critical events that enhance the catalytic efficiency of CDO.
Previous mechanisms for CDO propose that ferrous iron is primed
in the active site for the activation of dioxygen [9,11]. However, the
identification of Fe(III) in active crosslinked CDO suggests that fer-
ric iron activates the cysteine–thiolate ligand first prior to reaction
with dioxygen, analogous to the intradiol-cleaving catechol dioxy-
genases that utilize Fe(III) to activate the catecholate substrate.
Spectroscopic characterization of CDO by MCD spectroscopy
showed that cysteine binds directly to high spin Fe(III) in a similar
manner as it binds Fe(II) via its S atom [25]. The catalytic cycle in
CDO therefore likely proceeds with an oxidized iron center as out-
lined in Scheme 3 with cysteine coordinating Fe(III). Binding of cys-
teine to the Fe(III) center increases the affinity of the metal center
for dioxygen binding at an adjacent coordination site [26]. Coordi-
nation of cysteine to Fe(III) (Scheme 3, II) leads to electron transfer
from the thiolate to Fe(III) to form the thiyl radical and Fe(II). The
Fe(II) is then able to activate dioxygen forming an iron superoxo
intermediate (Scheme 3, III). Activation of the two substrates
would promote radical coupling between the thiyl sulfur and the
distal oxygen forming the cyclic iron peroxo complex (Scheme 3,
IV). Homolytic cleavage of the O–O bond would lead to the forma-
tion of a sulfoxy-cation and a single oxygen atom bound to the
metal center (Scheme 3, V). Recombination of the sulfoxy-cation
and the oxygen atom would generate the Fe(III) bound cysteine
sulfinic acid product which is subsequently released from the
active site by hydrolysis (Scheme 3, VI). The iron was not reduced
to the ferrous state when cysteine was added to crosslinked CDO
under either anaerobic or aerobic conditions. Therefore, it is likely
that electron transfer from the cysteine thiolate occurs once
dioxygen coordinates the iron center forming a ternary complex
that promotes electron transfer. This mechanism differs from the
intradiol dioxygenase enzymes where dioxygen reacts directly
with the activated catechol substrate. For the intradiol dioxygenase
enzymes, the oxygen-surrogate NO is unable to coordinate the
metal center unless the iron center has been reduced [27–28].
However, the oxygen-surrogate KCN was shown to coordinate
the Fe(III) center of CDO in the presence of Cys, suggesting that
oxygen is able to coordinate the ferric redox form [18].
geometry is also associated with an ꢀ5-fold increase in the kcat/K
m
value over CDO containing both isoforms. It can be rationalized
that the active site of crosslinked CDO may be in an entatic state
in which formation of the crosslink constrains the metal coordina-
tion environment to adopt a geometry which closely resembles the
reactive intermediate species formed during the course of the reac-
tion [23]. This would enhance the activity of CDO by lowering the
energy required to reorganize the active site during catalysis [23].
A similar mechanistic feature is observed in blue copper proteins in
which Cu(I) geometry is imposed on the Cu(II) active site which
enhances electron transfer [24]. Since most of the proposed mech-
anisms for cysteine substrate oxidation by CDO posits the initial
formation of Fe(III) superoxo species upon binding of dioxygen to
the enzyme–substrate complex, the results suggest that the active
site of crosslinked CDO with a ferric iron structurally resembles the
proposed Cys-bound Fe(III) intermediate. The substrate therefore
enters a preorganized active site, which would explain the
increased catalytic efficiency observed in crosslinked CDO.
As previously shown, iron, dioxygen, and cysteine are a require-
ment for crosslink formation [15–17]. Based on the redox proper-
ties of non crosslinked CDO and prior studies, a mechanism for
the formation of the Cys–Tyr crosslink is proposed (Scheme 2)
It was previously shown that CDO present in the cellular lysate
exists primarily in the non crosslinked ferrous form [19]. Exposure
of the enzyme to dioxygen during the purification process subse-
quently leads to formation of the Cys93–Tyr157 crosslink and oxi-
dation of the iron center. On the basis of this observation, it can be
inferred that within the cell, the predominant CDO isoform is non
crosslinked CDO and the crosslink is generated as a consequence of
exposing the enzyme to increasing concentrations of cysteine. In
[
11,15,17]. The cysteine substrate coordinates the active site iron
with the thiolate and amino groups as observed in the three-
dimensional structures of the CDO–cysteine complex [11]. Dioxy-
gen then reacts with Cys-bound Fe(II) CDO to form a Fe(III) super-
oxo species which is stabilized by the Tyr157 hydroxyl group
through H-bonding (Scheme 2, II). This is followed by proton cou-
pled electron transfer from Tyr157 to generate a tyrosyl radical

6
8
C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
Scheme 2. Proposed mechanism for crosslink formation.
Scheme 3. Proposed mechanism for cysteine oxidation.

C.W. Njeri, H.R. Ellis / Archives of Biochemistry and Biophysics 558 (2014) 61–69
69
rats, hepatic cysteine concentrations only vary between 0.02 and
[4] K. Yamaguchi, Y. Hosokawa, N. Kohashi, Y. Kori, S. Sakakibara, I. Ueda, J.
Biochem. 83 (1978) 479–491.
100 nmol/g and plasma cysteine concentrations vary between
.08 and 0.2 mM [29–30]. Such a tight regulation of free cysteine
[
[
5] J.B. Lombardini, T.P. Singer, P.D. Boyer, J. Biol. Chem. 244 (1969) 1172–1175.
6] M.H. Stipanuk, I. Ueki, J. Inherit. Metab. Dis. 34 (2011) 17–32.
0
is necessary because cysteine may be both cytotoxic and neuro-
toxic when present at high concentrations [31–33]. Previous stud-
ies demonstrated an increase in hepatic CDO activity in rats fed
high protein diets or diets supplemented with sulfur amino acids,
and in isolated hepatocytes incubated with 0.2 mM cysteine
[7] M.H. Stipanuk, C.R. Simmons, P.A. Karplus, J.E. Dominy Jr., Amino Acids 41
2011) 91–102.
(
[
8] C.R. Simmons, K. Krishnamoorthy, S.L. Granett, D.J. Schuller, J.E. Dominy Jr., T.P.
Begley, M.H. Stipanuk, P.A. Karplus, Biochemistry 47 (2008) 11390–11392.
[9] J.G. McCoy, L.J. Bailey, E. Bitto, C.A. Bingman, D.J. Aceti, B.G. Fox, G.N. Phillips Jr.,
Proc. Natl. Acad. Sci. U.S.A. 103 (2006) 3084–3089.
10] C.R. Simmons, Q. Liu, Q. Huang, Q. Hao, T.P. Begley, P.A. Karplus, M.H. Stipanuk,
J. Biol. Chem. 281 (2006) 18723–18733.
[
[
29,34–35]. The increase in activity was associated with an
increase in the lower molecular weight isoform, but the Cys–Tyr
crosslink had not been identified at that time. Thus, formation of
the crosslink appears to be a regulatory mechanism, which
enhances the catalytic efficiency of CDO when the cell is exposed
to increasing concentrations of cysteine. Differentiation between
crosslink formation or catalysis lies in the oxidation state of the
bound iron. The presence of Fe(II) in non crosslinked CDO favors
formation of the crosslink while the presence of Fe(III) in cross-
linked CDO favors oxidation of cysteine to cysteine sulfinic acid.
The observation that complete crosslink formation is achieved in
an enzyme with substoichiometric amounts of iron bound has
interesting mechanistic implications. Further investigations are
planned to investigate how the iron free active sites acquire the
crosslink.
[11] S. Ye, X. Wu, L. Wei, D. Tang, P. Sun, M. Bartlam, Z. Rao, J. Biol. Chem. 282
2007) 3391–3402.
(
[
12] M.S. Rogers, R. Hurtado-Guerrero, S.J. Firbank, M.A. Halcrow, D.M. Dooley, S.E.
Phillips, P.F. Knowles, M.J. McPherson, Biochemistry 47 (2008) 10428–10439.
[13] M.M. Whittaker, P.J. Kersten, N. Nakamura, J. Sanders-Loehr, E.S. Schweizer,
J.W. Whittaker, J. Biol. Chem. 271 (1996) 681–687.
[
14] R. Schnell, T. Sandalova, U. Hellman, Y. Lindqvist, G. Schneider, J. Biol. Chem.
80 (2005) 27319–27328.
2
[15] J.E. Dominy Jr., J. Hwang, S. Guo, L.L. Hirschberger, S. Zhang, M.H. Stipanuk, J.
Biol. Chem. 283 (2008) 12188–12201.
[
[
[
[
16] T. Kleffmann, S.A. Jongkees, G. Fairweather, S.M. Wilbanks, G.N. Jameson, J.
Biol. Inorg. Chem. 14 (2009) 913–921.
17] E. Siakkou, M.T. Rutledge, S.M. Wilbanks, G.N. Jameson, Biochim. Biophys. Acta
1814 (2011) 2003–2009.
18] W. Li, E.J. Blaesi, M.D. Pecore, J.K. Crowell, B.S. Pierce, Biochemistry 52 (2013)
9104–9119.
19] E.M. Imsand, C.W. Njeri, H.R. Ellis, Arch. Biochem. Biophys. 521 (2012) 10–17.
[20] S.E. Parkin, S. Chen, B.A. Ley, L. Mangravite, D.E. Edmondson, B.H. Huynh, J.M.
Bollinger Jr., Biochemistry 37 (1998) 1124–1130.
[
[
21] J.A. Crawford, W. Li, B.S. Pierce, Biochemistry 50 (2011) 10241–10253.
22] S.C. Chai, A.A. Jerkins, J.J. Banik, I. Shalev, J.L. Pinkham, P.C. Uden, M.J. Maroney,
J. Biol. Chem. 280 (2005) 9865–9869.
Acknowledgments
[
[
[
[
23] G. Chaka, J.L. Sonnenberg, H.B. Schlegel, M.J. Heeg, G. Jaeger, T.J. Nelson, L.A.
Ochrymowycz, D.B. Rorabacher, J. Am. Chem. Soc. 129 (2007) 5217–5227.
24] W.B. Church, J.M. Guss, J.J. Potter, H.C. Freeman, J. Biol. Chem. 261 (1986) 234–
237.
The authors wish to thank Dr. Martin Bollinger (Penn State Uni-
versity) for the insightful discussion on iron free expression. We
would also like to thank Dr. Evert Duin (Auburn University) for
assistance with the EPR instrument and for the use of his anaerobic
tent. We would also like to thank Dr. Paul Cobine (Auburn Univer-
sity) for the use of his ICP-AES instrument. The authors thank Ryan
Hasemeier for valuable discussions regarding the manuscript.
25] J.D. Gardner, B.S. Pierce, B.G. Fox, T.C. Brunold, Biochemistry 49 (2010) 6033–
6041.
26] S. Fetzner, Appl. Environ. Microbiol. 78 (2012) 2505–2514.
[27] E.C. Wasinger, M.I. Davis, M.Y. Pau, A.M. Orville, J.M. Zaleski, B. Hedman, J.D.
Lipscomb, K.O. Hodgson, E.I. Solomon, Inorg. Chem. 42 (2003) 365–376.
[
[
28] A.M. Orville, J.D. Lipscomb, J. Biol. Chem. 268 (1993) 8596–8607.
29] J.I. Lee, M. Londono, L.L. Hirschberger, M.H. Stipanuk, J. Nutr. Biochem. 15
(2004) 112–122.
Appendix A. Supplementary data
[
[
30] M.H. Stipanuk, J.E. Dominy Jr., J.I. Lee, R.M. Coloso, J. Nutr. 136 (2006) 1652S–
1659S.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.abb.2014.06.001.
31] P. Andine, O. Orwar, I. Jacobson, M. Sandberg, H. Hagberg, J. Neurochem. 57
(1991) 230–236.
32] A. Lehmann, Neuroscience 22 (1987) 573–578.
[
[
33] A. Lehmann, H. Hagberg, O. Orwar, M. Sandberg, Eur. J. Neurosci. 5 (1993)
References
1398–1412.
[
34] D.L. Bella, L.L. Hirschberger, Y. Hosokawa, M.H. Stipanuk, Am. J. Physiol. 276
(1999) E326–E335.
[
[
[
1] L. Ewetz, B. Sorbo, Biochim. Biophys. Acta 128 (1966) 296–305.
2] B. Soerbo, L. Ewetz, Biochem. Biophys. Res. Commun. 18 (1965) 359–363.
3] A. Wainer, Biochim. Biophys. Acta 104 (1965) 405–412.
[35] D.L. Bella, C. Hahn, M.H. Stipanuk, Am. J. Physiol. 277 (1999) E144–E153.