104
Y.H. Ouellet et al. / Journal of Inorganic Biochemistry 154 (2016) 103–113
(ChuV) [10]. Once inside the cytoplasm, the heme may be bound by
chaperones and transferred to other proteins or it may be metabolized
to release the sequestered iron. This later reaction may be carried out
by a heme oxygenase (HO) in Gram-negative bacteria that possess the
gene encoding this enzyme. Like the HO from higher organisms, bacte-
rial HO uses heme both as a substrate and as a cofactor to catalyze the
opening of the porphyrin ring thus producing biliverdin and carbon
monoxide and liberating the iron [11–15].
within ChuS reacts with H2O2 in the low micromolar range and we pro-
vide evidence on ferric (not ferrous) iron being released upon cleavage
of the heme by a mechanism that differs from that of the canonical HOs.
2. Material and methods
2.1. Chemicals and reagents
The reaction carried out by HO occurs at the heme-bound active site
and requires electrons, which are provided by cytochrome P450 reduc-
tase/NADPH or ascorbate in vitro, and molecular oxygen [12,13,16,17].
The complete reaction cycle requires 7 electrons in total and generates
carbon monoxide, ferrous iron and biliverdin. The mechanism of the
HO reaction has been elucidated and the detailed picture of the reactive
intermediates of the catalytic reaction emerged from biochemical stud-
ies in heme oxygenase 1 (HO-1) and a bacterial HO (HmuO), using var-
ious spectroscopy techniques (reviewed in [11–14,16]). In the HO
reaction, after the heme is first reduced and binds oxygen, the heme-
bound oxygen is activated by a second electron and a proton to form a
hydroperoxy complex (FeIII-OOH) from which hydroxylation of the
heme occurs and meso-hydroxyheme is formed. The initial oxygenated
complex of HO was characterized by resonance Raman spectroscopy
[18,19] and X-ray crystallography [19] which revealed an unusually
highly-bent geometry of the heme-bound O2. The short-lived
hydroperoxy complex was formed by cryoreduction from the oxidized
enzyme and was characterized by EPR and ENDOR spectroscopies
[20]. Notably, this intermediate was shown to be competent to hydrox-
ylate the heme to produce meso-hydroxyheme [20]. The details about
the coordination and electronic structure of the meso-hydroxyheme in-
termediate, which is stable in absence of molecular oxygen and could
exist in three equilibrium forms, was probed by EPR and resonance
Raman spectroscopies [21,22]. The subsequent intermediate,
verdoheme, which is formed spontaneously by the reaction of meso-
hydroxyheme with molecular oxygen, was probed with NMR [23] and
resonance Raman spectroscopies [21] and shown to be in a 6-
coordinate state. The complex of verdoheme with HO-1 was also char-
acterized by X-ray crystallography to get a better picture of the mecha-
nism for the opening of the porphyrin ring to the final product biliverdin
[24,25].The enzymatic degradation of heme by mammalian HOs occurs
quite selectively at the alpha-meso-carbon. On the contrary, the chem-
ical oxidation of heme in the presence of excess reductant and molecu-
lar oxygen, termed coupled oxidation, is rather unspecific attacking
essentially all four meso-carbons [26,27].
Hemin, catalase, sodium ascorbate, ABTS, guaiacol and horseradish
peroxidase (HRP) were from Sigma. Biliverdin-hydrochloride was
from Frontier Scientific (Newark, DE).
2.2. ChuS expression and purification
The expression and purification of the recombinant apo-ChuS, with a
N-terminal tag composed of 6 histidine residues, was performed in ac-
cordance with our previously published method [36]. As a control to
verify that the histidine tag was not involved in the heme degradation
activity of ChuS, we also expressed and purified the ChuS protein with-
out any tag (Supp. material). The histidine-tagged and untagged pro-
teins degraded heme similarly using ascorbate (Fig. 1 and Supp.
Fig. 1). All measurements reported in this study were obtained with
the histidine-tagged protein for which the crystal structure is known
[36,37]. Protein purity was verified by SDS-PAGE. As purified, the ChuS
protein is devoid of heme but displays a light purple color presumably
because the enzyme is active inside the bacteria and a fraction of the
protein contains a porphyrin degradation product [36]. This product
can be extracted using butanol as a solvent (see Section 2.8). Protein
concentration was quantified by the Bradford method.
2.3. Preparation of heme-bound ChuS (Holo-ChuS)
Holo-ChuS was prepared by adding a six-fold molar excess of hemin
(solubilized in 15 mM NaOH and neutralized in 50 mM Tris–HCl pH 8)
to the apo-protein solution and incubated 15–30 min at room tempera-
ture. Removal of excess heme was performed by loading the heme-
protein mixture onto a 10 ml DEAE Sepharose Fast Flow column (GE
Healthcare) equilibrated with 50 mM Tris–HCl pH 8. After washing
with the same buffer, the holo-protein complexes was eluted with
50 mM Tris–HCl pH 8 buffer containing 250 mM NaCl. The collected
fractions were dialyzed (Spectrapor membranes, 10 kDa MWCO)
against 0.1 M sodium phosphate pH 7 overnight at 4 °C. The dialysate
was concentrated using a 10 kDa cut-off Centricon (Millipore) and
kept at −80 °C. The protein and heme contents were quantified by
the Bradford (Bio-Rad Laboratories) and pyridine hemochrome
methods, respectively [38]. We determined that holo-ChuS contains
one heme per monomer of protein (1.1 0.1). All protein concentra-
tions throughout this paper are reported based on the heme content.
All experiments were performed in 0.1 M sodium phosphate buffer at
pH 7.0.
Bacterial HOs similar to those of mammalian, termed canonical HOs,
are found in several pathogens including Corynebacterium diphtheria
[19,28,29], Neisseria meningitidis [30,31] and Pseudomonas aeruginosa
[32]. In Gram-positive bacteria, the IsdG family of proteins performs a
similar function to HOs although they are not structurally related and
yield porphyrin products different form biliverdin [33–35]. Notably, a
HO is apparently absent in E. coli strains, including O157:H7, based on
genomic data mining. Thus, the fate of the heme once it reaches the cy-
toplasm is not clearly understood in E. coli O157:H7 or in bacteria that
have no homologs of canonical HOs or IsdG [3].
2.4. UV–vis characterization of heme degradation
We previously reported that the ChuS protein of E. coli O157:H7, a
protein encoded by the chuS gene of the chu gene cluster, is able to
bind and degrade heme in a reaction that releases carbon monoxide
[36]. The physiological role for the ChuS protein in iron acquisition re-
mains however to be established [3]. Since this protein has no structural
homology to canonical HOs neither to the IsdG family of proteins [36,
37], we have pursued a thorough investigation of its heme degradation
mechanism. In this work, we present a detailed analysis of the heme
degradation reaction by ChuS using stopped-flow optical absorption
spectroscopy and EPR spectroscopy and we identified reaction products
by NMR spectroscopy and mass spectrometry (MS) in order to gain fur-
ther knowledge about the role of proteins of the heme acquisition path-
way of E. coli O157:H7 and of other bacteria. We show that the heme
The characterization of ChuS heme degradation reaction was studied
at 20 °C with either sodium ascorbate or with hydrogen peroxide on a
Cary 3E spectrophotometer (Varian) equipped with a temperature-
controlled multicell holder. Sodium ascorbate was added to a 10 μM so-
lution of holo-ChuS to a final concentration of 100 μM. The absorption
spectra of ChuS upon addition of sodium ascorbate were recorded at 5
minute intervals during 60 min. Where indicated, catalase (50 to
300 U) was added to the reaction mixtures. For the peroxide-driven re-
action, hydrogen peroxide (final concentrations of 5–20 μM) was added
to a 10 μM holo-ChuS. Heme degradation was also followed by acquiring
the optical absorption spectra at 5 minute intervals. The concentration
of the H2O2 stock solution was determined using an absorption coeffi-
cient of 43.6 M−1 cm−1 at 240 nm [39]. To obtain the spectra of the