Crystal Structure of the Dioxygen-bound Heme Oxygenase from

1

Crystal Structure of the Dioxygen-bound Heme Oxygenase from Corynebacterium

diphtheriae: Implications for Heme Oxygenase Function

by

Masaki Unno&, Toshitaka Matsui&, Grace C. ChuH,a, Manon Couture',b,

Tadashi Yoshida†, Denis L. Rousseau', John. S. Olson††, and Masao Ikeda-Saito&,H

From the &Institute of Multidisciplinary Research for Advanced Materials, Tohoku University,

Katahira, Aoba, Sendai 980-8577, Japan, the HDepartment of Physiology and Biophysics, Case

Western Reserve University School of Medicine, Cleveland, Ohio 44106-4970, the 'Department

of Physiology and Biophysics, Albert Einstein College of Medicine, Bronx, New York, 10461, †Department of Biochemistry, Yamagata University School of Medicine, Yamagata 990-9585,

Japan, and the ††Department of Biochemistry and Cell Biology, Rice University, Houston, Texas

77005-1892

Running title: Crystal Structure of the Oxy C. diphtheriae Heme Oxygenase

Correspondences to:

Dr. Masaki Unno, e-mail: [email protected]

Dr. Masao Ikeda-Saito, e-mail: [email protected]

JBC Papers in Press. Published on February 13, 2004 as Manuscript M400491200

Copyright 2004 by The American Society for Biochemistry and Molecular Biology, Inc.

by guest on April 11, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

2

Footnotes

*Footnote to the Title: This work was supported in part by National Institutes of Health grants

GM57272 (MIS), GM54806 (DLR), GM35649 (JSO), and HL40720 (JSO), a grant C-612 from

the Robert A Welch Foundation (JSO), a grant from Takeda Science Foundation (MIS), and

grants-in-aid from the Ministry of Education, Science, Culture and Sports, Japan 12147201

(MIS), 14380300 (MIS), 14580641 (TY), 14740358 (TM), and 15770095 (MU). A part of the

work was conducted at the Hybrid Nano-Material Research Center of the Institute of

Multidisciplinary Research for Advanced Materials, Tohoku University.

The atomic coordinates and structure factors (code 1V8X) have been deposited in the

Protein Data Bank, Research Collaboratory for Structural Bioinformatics, Rutgers University,

New Brunswick, NJ (http://www.rcsb.org/).

Present addresses:

a Amgen, Inc., Thousand Oaks, CA. 91320 b Department of Biochemistry and Microbiology, University of Laval Québec, Québec Canada G1K 7P4

Footnotes to the text

1. Abbreviations used are: HO, heme oxygenase; Mb, myoglobin; Hb, hemoglobin; r.m.s., root

mean square.

2. Irradiation of oxy HmuO by γ-ray at cryogenic temperatures shows that one-electron

reduction of the oxy HmuO yields the catalytically active ferric hydroperoxo species, which

self-hydroxylates to form the ferric α-meso-hydroxyheme intermediate, as has been

established for HO-1 (R. M. Davydov, G. C. Chu, M. Ikeda-Saito, and B. M. Hoffman,

manuscript in preparation).

3. T. Matsui, M. Unno, and M. Ikeda-Saito, unpublished results


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

3

Summary

HmuO, a heme oxygenase of Corynebacterium diphtheriae, catalyzes degradation of heme using

the same mechanism as the mammalian enzyme. The oxy form of HmuO, the precursor of the

catalytically active ferric hydroperoxo species, has been characterized by ligand binding kinetics,

resonance Raman spectroscopy, and X-ray crystallography. The oxygen association and

dissociation rate constants are 5 µM-1 s-1 and 0.22 s-1, respectively, yielding an O2 affinity of 21

µM-1 that is ~20 times greater than that of mammalian myoglobins. However, the affinity of

HmuO for CO is only 3 to 4-fold greater than that for mammalian myoglobins, implying the

presence of strong hydrogen bonding interactions in the distal pocket of HmuO that

preferentially favor O2 binding. Resonance Raman spectra show that the Fe-O2 vibrations are

tightly coupled to porphyrin vibrations, indicating a highly bent Fe-O-O geometry that is

characteristic of the oxy forms of heme oxygenases. In the crystal structure of the oxy form, the

Fe-O-O angle is 110º; the O-O bond is pointed toward the heme α-meso-carbon by direct steric

interactions with Gly135 and Gly139; and hydrogen bonds occur between the bound O2 and the

amide nitrogen of Gly139 and a distal pocket water molecule, which is a part of an extended

hydrogen bonding network that provides the solvent protons required for oxygen activation. In

addition the O-O bond is orthogonal to the plane of the proximal imidazole side chain, which

facilitates hydroxylation of the porphyrin α-meso-carbon by preventing premature O-O bond

cleavage.

(245 words)


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

4

INTRODUCTION

Heme oxygenase, HO1, catalyzes regioselective oxidative conversion of iron

protoporphyrin IX (heme) to biliverdin IX, iron, and CO (1). HO is not a hemeprotein by itself

but utilizes heme as both a prosthetic group and a substrate. In mammalian systems, where

electrons are supplied by NADPH through NADPH-cytochrome P450 reductase (2), HO is the

enzyme responsible for excess heme degradation (1) and iron recycling (3). The product CO has

been implicated as a messenger molecule in various physiological functions (4, 5). Although

most of the structural and functional studies have been conducted on the soluble, truncated form

of isoform-1 of mammalian heme oxygenase, HO-1 (6-8), the enzyme is also present in some

pathogenic bacteria where it is essential for heme-based iron acquisition from a host lacking in

free extracellular iron (9-11). Two HO proteins from pathogenic bacteria have been

characterized in some detail: HmuO from Corynebacterium diphtheriae, and HemO from

Neisseria meningitidis (12-16). In comparison to the mammalian HO, neither of them is

membrane-bound. Instead, they are soluble and have smaller molecular masses: HmuO, 24 kDa,

and HemO, 26 kDa. HmuO has 33% sequence identity to the first 221 amino acids of human

HO-1. Despite their differences in size, the two bacterial HO proteins have overall protein folds,

heme environments, and catalytic mechanisms very similar to those for mammalian HO-1 (12-

16).

In its catalytic cycle (Figure 1), HO first binds one equivalent of heme to form a ferric

heme-HO complex. The first electron donated from the reducing substrate converts the heme

iron to the ferrous state. Then O2 binds to reduced 5-coordinate heme to form a meta-stable oxy

complex. One-electron reduction of the oxy form generates a ferric hydroperoxo complex,

which self-hydroxylates the α-meso-carbon of the porphyrin ring (6, 17). The latter reaction is

different from what occurs in P450 enzymes, in which the O-O bond of the hydroperoxo

complex is heterolytically cleaved to generate an actively hydroxylating, ferryl (Fe4+=O)

intermediate (18). Ferric α-meso-hydroxyheme is then converted to biliverdin by multiple

oxidoreductive steps involving a verdoheme intermediate (6, 12, 19).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

5

The oxy form of the heme complex of heme oxygenase has several unique features

relevant to HO catalytic function (20-22). O2 binds to the complex with high oxygen affinity

due to a very small oxygen dissociation rate constant (21). This is one of the reasons why HO

catalysis is not severely inhibited by its product CO in air-saturated buffer. The cause for the

small O2 dissociation rate is assumed to involve hydrogen bonding between the bound ligand

and its adjacent protein environment but, until this report, has not been confirmed by direct

structure determination (21-23). Analysis of the Fe-O2 vibration modes predicts that the Fe-O-O

unit is highly bent by steric interactions between the bound O2 and the residues in the distal

pocket (20). These results indicating both steric restriction and hydrogen bonding led to the

proposal that the terminal oxygen is “held” in a position adjacent to the α-meso-carbon of the

porphyrin ring, facilitating a highly regiospecific oxygenation (14, 20, 24), but again this

orientation has not been confirmed directly by X-ray crystallography. A hydrogen bond with an

adjacent exchangeable proton from a distal pocket water molecule has also been proposed (13,

22), and this proton is thought to become the hydroperoxo proton after one-electron reduction of

the oxy form (17, 25). This hydrogen bonding proposal was based on EPR studies with HmuO

and HO containing oxy cobalt porphyrin and also requires direct structural confirmation.

Until this report, a crystal structure of the oxy form of heme oxygenase had not been

reported due to the meta-stable nature of the Fe-O2 intermediate. The ferrous NO complex has

been used as a mimic of the oxy form in recent structural studies, and has provided useful

information about the active site of HO (23, 26). However, the structure of the oxy complex is

needed to more accurately delineate the mode of oxygen binding. The heme oxygenase from C.

diphtheriae (HmuO) has been thoroughly characterized structurally and enzymatically (12-14),

and we have now been able to obtain crystals of oxy form of this enzyme. Oxygenated crystals

were obtained by using strictly anaerobic conditions for reduction of ferric crystals, carefully

removing excess reducing agent, exposing the one-electron reduced sample to mother liquor

equilibrated with 1 atm of O2 for 10 minutes, and then flash freezing the final oxygenated

samples. These crystals were used to determine the structure of oxy HmuO to 1.85 Å resolution.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6

We have also examined O2 and CO binding to the ferrous heme-HmuO complex and

characterized the resonance Raman spectrum of the oxy complex. The crystal structure shows

that bound oxygen is tightly confined by two Gly residues, which direct the O-O axis toward the

heme α-meso-carbon atom. Bound O2 also forms a hydrogen bond with a solvent accessible

distal pocket water molecule that both slows oxygen dissociation and provides protons for the

formation of the hydroperoxo intermediate, which hydroxylates the porphyrin ring.

EXPERIMENTAL PROCEDURES

Expression, purification, and reconstitution of the recombinant HmuO with heme were

carried out as described previously (12). Ligand binding reactions were measured using stopped-

flow and flash photolysis techniques as described previously (27). The CO form of the heme-

HmuO complex was prepared by diluting the ferrous forms of the complex into the anaerobic

buffer solutions containing known CO concentrations. The oxygenated form was made as

follows. First, the ferric heme-HmuO complex was reduced by sodium dithionite in

deoxygenated buffer. Then, the reduced form of the complex was loaded on to a column of

Sephadex G-25 and eluted with buffer containing the known oxygen concentrations. Formation

of the complex and sample integrity were confirmed by the optical absorption spectra recorded

before and after the flash photolysis measurements. All the measurements were carried out in

0.1 M phosphate buffer, pH 7 at 20o C.

Resonance Raman spectra of the oxy form were obtained by the use of a continuous flow

rapid mixing apparatus as described previously (28, 29). Continuous laser illumination caused

oxidation of the oxy form of the heme-HmuO complex, so only the flow method yielded usable

Raman spectra of the oxy form of HmuO. The deoxy form of the heme-HmuO complex was

mixed with either 16O2 or 18O2 saturated buffer (0.1 M phosphate pH 7 containing 1 mM EDTA)

with a flow rate of 1.2 µs/µm. The Raman data were taken at 25 ms after mixing at 4o C with

excitation by the 413.1 nm line of a Krypton laser (40 mW). The Raman shifts were calibrated


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

7

by using indene as standard. Optical absorption spectra of the collected sample were taken after

Raman measurements to confirm formation of the O2-bound heme-HmuO.

The crystals of the oxy form of the heme-HmuO complex were prepared as follows. The

crystals of the ferric heme-HmuO complex were obtained by a hanging drop vapor diffusion

method as described previously (14). The ferric HmuO crystals were transferred into an N2-

bubbled reservoir solution containing 50 mM MES, pH 6.1, 2.2 M ammonium sulfate, and 5%

(w/v) sucrose. The crystals were brought into a nitrogen glove box, where they were transferred

into the reservoir solution containing 50 mM sodium dithionite and 5% sucrose. The crystals

were soaked for 10 minutes to ensure complete conversion to the ferrous form. In order to

remove dithionite and its oxidized products, the ferrous crystals were then transferred into

another reservoir solution containing 15% sucrose without dithionite and were washed by

repeated pumping with a micropipette. The crystals were further transferred into another

reservoir solution that contained 20% sucrose without dithionite and washed as before. After

washing, the crystals were taken out of the globe box and were transferred into the reservoir

solution containing 25% sucrose, which had been equilibrated thoroughly with 1 atm of pure O2

prior to soaking the crystals. After soaking for ten minutes, the crystals were picked up by nylon

loops and flash frozen with liquid N2. Formation of the oxy complex was confirmed by

recording light absorption spectra of the crystals by a microspectrophotometer (4DX Systems

AB, Uppsala, Sweden) in the visible region. During the absorption measurements, the crystals

were kept at 100 K by an Oxford liquid nitrogen flow cryostat. The crystals used for the X-ray

diffraction experiments were mostly in the oxy form with the remainder being the ferric form (~5

to 10%).

X-ray diffraction data sets were collected with a Rigaku R-Axis V imaging-plate using

1.0 Å synchrotron radiation at BL41XU of SPring-8. The temperature around the crystals was

maintained at 100 K throughout data collection. The oscillation angle, camera length and

exposure time were 1 degree, 200 mm and 5 seconds, respectively. The incident X-ray intensity

was attenuated by use of a 0.6 mm thick aluminum film to reduce radiation damage to the


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

8

crystals. Data were integrated, merged, and processed with HKL-2000 (30). Diffraction data

with 100% completeness was obtained from 180 frames. The crystals belonged to the space

group P21 with unit-cell parameters of a = 54.0 Å, b = 63.0 Å, c = 107.5 Å, and β = 101.0º and

contained three HmuO molecules per one asymmetric unit. Diffraction statistics are summarized

in Table I.

Rigid-body refinement up to 3.0 Å was performed by CNS (31) using the structure of

ferrous HmuO (14, PDB code 1IW1) as a starting model. The phases were extended to 1.85 Å

by density modifications including solvent flattening, non-crystallographic symmetry (NCS)

averaging between three molecules, and histogram matching with program DM from the CCP4

suites (32). The structure was refined to fit to the observed structure factors using simulated

annealing (T = 4000 K) for the 40 - 1.85 Å resolution data by CNS. The crystallographic

refinements with simulated annealing and individual B-factor refinement with CNS were

performed to calculate an unbiased model. At this stage, NCS restraint was removed, because

each of the three molecules in the asymmetric unit had slightly different conformations, as was

the case for the ferric and ferrous heme-complexes of HmuO (14). After several refinement

cycles, water molecules were added to the model by CNS and were moved and/or removed by

manual inspection in both of σA-weighted 2Fo-Fc and Fo-Fc maps with the program O (33). The

model was further refined using the maximum-likelihood target with the program REFMAC5

(34). The O2 ligand was modeled into a rugby ball-shaped density next to the iron at final stages

of the refinement. After introduction of alternative conformations for several residues, and the

translation-liberation-screw (TLS) refinement (35), the final R and Rfree factors dropped to 0.153

and 0.192, respectively. The final refinement statistics are summarized in Table I. Throughout

the model building and refinement process, 10% of the reflections were excluded to monitor the

Rfree value. Several residues located N- and C-terminal regions were not visible in the electron

density map probably due to their disorder. Drawings were made by MOLSCRIPT (36), and

BOBSCRIPT (37).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

9

RESULTS

Association and dissociation rate constants for O2 binding to the ferrous form of the

heme-HmuO complex are compared to those of rat HO-1, and wild-type and V68N pig Mb in

Table II. Bimolecular rebinding of O2 to the heme-HmuO complex after flash photolysis is

monophasic, and the observed rate depends linearly on the O2 concentration. The association

rate constant for O2 binding to the heme-HmuO complex, k'O2 = 5 µM-1 s-1, is 3-fold smaller

than that of mammalian Mbs (Table II) (38). However, the O2 dissociation rate constant (kO2)

for the oxy heme-HmuO complex is ~0.22 s-1, which is ~70 times smaller than that of Mb (~15

s-1). The oxygen equilibrium constant (KO2) of HmuO is 21 µM-1, which is ~20 times greater

than that of Mb

As shown in Table II, the O2 and CO binding parameters of the heme-HmuO complex

are very similar to those of the HO-1 complex (21). Both heme oxygenases have O2 association

equilibrium constants that are 20 to 30-fold greater than those for mammalian Mbs, but the HO-1

and HmuO affinities for CO are only 2.5 to 4-fold greater than those for Mbs. As result, the

heme oxygenases show very low M-values (KCO/KO2 ≈ 5). The high KO2 and low M values are

reminiscent of mutant and naturally occurring heme proteins with multiple hydrogen bonds to

bound dioxygen (39-41) (V68N pig Mb and Ascaris HbII in Table II). In these proteins, the

distal pocket, hydrogen bond donors sterically restrict access to the iron atom. In the case of O2

binding, steric hindrance is overcome by the formation of much more favorable electrostatic

interactions that strongly and preferentially stabilize the highly polar Fe-O2 complex.

Traces A and B in Figure 2 show the low frequency region resonance Raman spectra for

the 16O2 or 18O2-bound forms of the heme-HmuO complex. Their difference spectrum is shown

in trace C, which reveals several oxygen isotope sensitive Raman lines. The largest isotope shift

(24 cm-1) is in the line at 565 cm-1 in the difference spectrum. The oxygen isotope difference

spectrum below 450 cm-1 is complicated: the major lines at 415 and 351 cm-1, which are as

intense as the 565 cm-1 line in the difference spectrum, have shifts of 11 and 8 cm-1; other


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

10

weaker shifts are also observed. These isotope shifts are essentially the same as those detected

for the oxygen isotope difference spectrum of the rat HO-1 complex shown as trace D (20). On

the basis of the Raman line assignment established for the rat HO-1 complex (20), the line at 565

cm-1 is assigned as the Fe-O2 stretching (νFe-O2) mode. The line at 415 cm-1 is assigned to the

Fe-O-O bending (δFe-O-O) mode, but it is strongly coupled with many porphyrin modes. Because

of this strong coupling, the isotope shift of the line itself is very small and the shift is distributed

over the many porphyrin modes in the 280 to 400 cm-1 region, as described for the rat HO-1

complex (20). The close similarity of the oxygen isotope difference spectra between the HO-1

and HmuO complexes shows that O2 binds to the heme group of the HmuO complex with a

bending angle of ~110o, as estimated for the rat HO-1 complex.

Initial attempts were made to generate oxy HmuO crystals by immersing the dithionite-

reduced deoxy crystals of the heme HmuO complex directly into an O2-saturated reservoir

solution, as has been done for Mb (42). This method, however, resulted in incomplete

oxygenation, auto-oxidation, and partial verdoheme formation in the HmuO crystals, as judged

by the light absorption spectra. Removal of excess dithionite and its oxidation products from the

deoxy crystals in an inert nitrogen glove box made it possible to obtain HmuO crystals with a

high content of the oxy form and without appreciable formation of ferric heme and verdoheme

(Supplemental data).

Since radiolytic irradiation at cryogenic temperatures readily converts oxy HO to the

ferric hydroperoxo form (17, 43)2, this intermediate might have been generated during the X-ray

diffraction measurements at SPring-8. To assess this possibility, the light absorption spectra of

the crystals were also measured after the X-ray diffraction measurements. Since the spectrum of

the ferric hydroperoxo intermediate is different from that of the ferrous oxy HO species (43), its

formation would be discernable. However, the spectra of the crystals after the diffraction

measurements were essentially the same as those taken before the measurements (Supplemental


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

11

data), indicating that ferric hydroperoxo species was not generated during the diffraction

measurements.

The structure of the ligand binding site in the oxy form of the heme-HmuO complex is

shown in Figure 3. Bound O2 is found at almost full occupancy in all three HmuO molecules in

the asymmetric unit. The structure of the ligand-free ferrous heme-HmuO complex (14) is little

affected by O2 binding. The root mean square displacements of the Cα atoms between the

ligand-free and the oxy forms are small and equal to 0.182, 0.190, and 0.163 Å for the three

HmuO molecules in the asymmetric unit. In agreement with the resonance Raman results, O2

binds end-on to the heme iron with a bend angle of ~110o, which is more acute than the bend

angle reported for the high resolution crystal structures of oxy forms of Mb (122o) (44),

Scarpharca dimeric HbI (155o and 135o) (45), cytochrome P-450CAM (142o) (46), and

horseradish peroxidase (126o) (47). In HmuO the distance between the terminal oxygen atom of

bound O2 and the porphyrin α-meso-carbon atom is ~3.4 Å. The bond length between the heme

iron and the first oxygen atom of the bound ligand is ~1.93 Å, which is slightly longer than those

reported for the oxy forms of most other heme proteins. The proximal His Nε–Fe bond is ~2.13

Å, which is similar to those found in other oxyheme proteins, and there is practically no

displacement of the iron atom from the mean plain of the porphyrin ring.

Gly135 and Gly139 are tightly packed around bound O2 (Figure 4). The Cα atom of

Gly139 is ~3 Å from the center of the O-O bond. The amide nitrogen of Gly139 is located

within hydrogen bonding distance to the oxygen atom bound to the heme iron (~3 Å). The

carbonyl oxygen of Gly135 is also very close to both oxygen atoms (~3 Å). This tight packing

by the two Gly residues points bound O2 directly toward the α-meso-carbon atom of the

porphyrin ring and severely restricts any rotation about the Fe-O2 bond.

The terminal oxygen atom also forms a strong hydrogen bond with an adjacent distal

pocket water molecule (Wat1), which is located ~2.7 Å from the second oxygen atom. Wat1 is

part of a hydrogen bonding lattice, which includes three other internal water molecules (Wat2,


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

12

Wat3, Wat4) and the amide and carboxylate side chains of Gln 46 and Asp136, respectively

(Figure 3). Asp136, which is equivalent to the catalytically critical Asp140 of HO-1 (25, 48, 49),

stabilizes Wat1 and Wat2. Wat2 is connected by Wat3 and Wat4 to the side chain of Gln46,

which, in turn, is connected by another water molecule to the surface amino acid, Asp86. Thus,

there is a hydrogen bonding network connecting bound O2 all the way out to the solvent water

phase, as found previously in the ferric and ligand-free ferrous forms of the heme-HmuO

complex (14).

DISCUSSION

HmuO, like HO-1, yields only the α-isomer of biliverdin (12). Regioselectivity of HO

catalysis is defined at the first oxygenation step (50), where the remote hydroperoxo oxygen

attacks the porphyrin α-meso-carbon to form α-meso-hydroxyheme (6, 25). Previous studies

have suggested that regiospecific degradation of heme by HO activity is primarily under the

steric control. The kinked distal helix lies directly above the heme plane and appears to

physically cover the β-, γ-, and δ-meso carbon atoms of the porphyrin ring, leaving only the α-

meso-carbon accessible by the hydroperoxo ligand (14, 16, 24, 26).

The crystal structure of the oxy form of the heme-HmuO complex confirms this view and

shows that the two distal Gly residues at the “kink” in the distal helix direct the geometry of the

Fe-O-O complex for regiospecific hydroxylation of the α-meso-carbon atom of the heme group

(Figure 4). A small cleft created by two distal Gly residues is present in the ligand-free ferrous

complex of HmuO (14), and O2 binds to the heme iron without significant rearrangement of this

distal pocket structure (Figure 3). The steric pressure imposed by these two Gly residues is

responsible for the acute Fe-O2 bending angle and the orientation of the O-O bond toward the α-

meso-carbon atom.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

13

In agreement with previous EPR studies of the oxy cobalt-porphyrin HO complex (13,

22), bound O2 forms a favorable hydrogen bond with a distal pocket water molecule (Wat1,

Figure 3) with a geometry and distance suitable for efficient proton transfer. This internal water

molecule is part of a hydrogen bonding network that extends from the Fe-O2 unit to solvent

water. This network provides a channel for efficient proton transfer to the active site, which is

required for formation of the ferric hydroperoxo species after one electron reduction of the oxy

form, activation of the O-O bond for cleavage, and hydroxylation of the α-meso-carbon (25, 48,

49). The second water molecule (Wat2) of the network is hydrogen bonded to Asp136.

Mutagenesis studies show that disruption of the hydrogen bonding network by replacing Asp136

by Phe results in the formation of the inactive ferryl species during HmuO catalysis3. Thus,

without proton transfer, an inactive ferryl species is generated instead of α-meso-hydroxyheme

(25, 48).

The proximal His (His20) in oxy HmuO retains a “staggered” conformation with respect

to the pyrrole nitrogen atoms, which is similar to that seen in high affinity Hbs, i.e.

leghemoglobin (51, 52) or Ascaris Hb (53). The imidazole plane of the proximal His side chain

is in-line with the meso β-δ carbon axis of the porphyrin ring, as in the ferrous ligand-free heme-

HmuO complex (14). Consequently, the projection of the dioxygen ligand onto the heme plane

is nearly perpendicular to the plane of the proximal imidazole base. In this configuration, the π-

orbitals of the proximal imidazole and the iron bound hydroperoxo species are orthogonal to the

iron dxz and dyz orbitals (54). This coordination structure decreases the “push-effect” of the

proximal His imidazole (14, 18) and stabilizes the O-O bond of the iron bound hydroperoxo,

inhibiting the formation of an inactive ferryl intermediate (7).

The O2 equilibrium constant of the heme-HmuO complex (~20 µM-1) is similar to those

of mammalian HO isoforms (30 - 80 µM-1) (21). In both the mammalian and bacterial HO cases,

high oxygen affinity is due primarily to a very small oxygen dissociation rate constant (~0.2 s-1).

The structure of oxy HmuO shown in Figure 3 provides a mechanistic interpretation of both the

high affinity and slow dissociation of O2. The staggered geometry of the coordinated imidazole


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

14

base (His20 side chain) in HumO and the pyrrole nitrogen atoms favors the binding of all ligands

(42); however, the sterically restricted distal pocket counteracts this favorable proximal effect.

As result, the affinity of HmuO for CO is only 3 to 4-fold greater than that of mammalian Mbs,

which have an unfavorable, eclipsed proximal geometry. However, in the case of HumO, two

strong hydrogen bonds preferentially stabilize the highly polar Feδ(+)-O-Oδ(-) complex. One

interaction is with the amide N-H of Gly139 and the other is with the distal pocket water

molecule, Wat1 (Figures 3 and 4).

Using mutagenesis, Wilkinson, Phillips, Olson, and their colleagues have demonstrated

unambiguously that increasing the strength and number of hydrogen bonds donated from the

distal pocket amino acids to bound O2 lowers the rate of oxygen dissociation from Mb (39, 41,

55, 56). The best example is V68N pig Mb, in which Val at the E11 helical position is replaced

with Asn. Two strong hydrogen bonds are seen in the crystal structure of V68N pig MbO2, one

from the distal His and the other from Asn at E11. As a result, kO2 decreases from 14 s-1 for

wild-type to 0.67 s-1 for the V68N mutant, and KO2 increases 4-fold. Naturally occurring

Ascaris suum Hb has an extremely low oxygen dissociation rate constant (kO2 = 0.004 s-1) due to

two strong hydrogen bonds between distal pocket Tyr and Gln side chains and bound O2 (53, 57).

For all three heme proteins, V68N Mb, Ascaris Hb, and HmuO the multiple hydrogen bonds in

the distal pocket preferentially stabilize bound O2 and not CO. As a result, the M-values

(KCO/KO2) for these proteins are very small, ≤ 10, compared to mammalian Hbs (M = 200 - 400),

Mbs (M = 30 - 40), and model hemes (M = 1,000 - 4,000) (Table II) (39, 58). In the case of

heme oxygenase, the low M-value is required to prevent inhibition of enzymatic activity by CO,

which is a product of the heme degradation reaction (21).

CONCLUSION - There are remarkably close similarities in the oxygen binding parameters, the

resonance Raman spectra, and distal pocket structures of HmuO and mammalian HO-1. Thus,

mammalian HO-1 and bacterial heme oxygenase, HmuO, almost certainly share a common

mechanism for facilitating O2 binding, pointing the bound ligand toward the α-meso-carbon

atom of the porphyrin ring, and providing protons for the hydroxylation reaction. The structure


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

15

of the oxy form of HmuO provides direct proof of the key roles played by two highly conserved

distal pocket Gly residues in determining regioselectivity in both types of heme oxygenases. In

addition, the HmuO structure also confirms the importance of hydrogen bonding interactions

with distal pocket water molecules for preferentially stabilizing bound O2 and for providing

protons for the formation of the activated hydroperoxo intermediate.

Acknowledgement

We thank members of the BL41XU Structural Biology Beam Line team for help in data

collection at SPring-8, and Dr. Sakabe’s research team at the Photon Factory for the loan of the

Oxford liquid nitrogen flow cryostat.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

16

REFERENCES

1. Maines, M. D. (1997) Annu. Rev. Pharmacol. Toxicol. 37, 517-554

2. Schacter, B. A., Nelson, E. B., Marver, H. S., and Masters, B. S. S. (1972) J. Biol. Chem.

247, 3601-3607

3. Poss, K. D., and Tonegawa, S. (1997) Proc. Natl. Acad. Sci. U.S.A. 94, 10919-10924

4. Dioum, E. J., Rutter, J., Tuckerman, J. R., Gonzalez, G., Gilles-Gonzalez, M.-A., and

McKnight, S. L. (2002) Science 298, 2385-2387

5. Suematsu, M., and Ishimura, Y. (2000) Hepatology 31, 3-6

6. Ortiz de Montellano, P. R. (2000) Curr. Opin. Chem. Biol. 4, 6388-6349

7. Ikeda-Saito, M., and Fujii, H. (2003) in “Paramagnetic Resonance of Metallobiomolecules”

(Tesler, J., ed.) ACS Symposium Series 858, 97-112, American Chemical Society,

Washington, D.C.

8. Colas, C., and Ortiz de Montellano, P. R. (2003) Chem. Rev. 103, 2305-2332

9. Wandersman, C., and Stojilijkovic, I. (2000) Curr. Opin. Micobiol. 3, 215-220

10. Genco, C., and Dixon, D. W. (2000) Mol. Microbiol. 39, 1-11

11. Schmitt, M. P. (1997) J. Bacteriol. 179, 838-845

12. Chu, G. C., Katakura, K., Zhang, X., Yoshida, T., and Ikeda-Saito, M. (1999) J. Biol. Chem.

274, 21319-21325

13. Chu, G. C., Tomita, T., Sonnichsen, F. D., Yoshida, T., and Ikeda-Saito, M. (1999) J. Biol.

Chem. 274, 24490-24496

14. Hirotsu, S., Chu, G. C., Unno, M., Lee, D.-S., Yoshida, T., Park, S.-Y., Shiro, Y., and

Ikeda-Saito, M. (2004) J. Biol. Chem. in press Published on Web on November 26, 2003 as

Manuscript M311631200


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

17

15. Zhu, W., Wilks, A., and Stijilijkovic, I. (2000) J. Bacteriol. 182, 6783-6790

16. Schuller, D. J., Zhu, W., Stijilijkovic, I., Wilks, A., and Poulos, T. L. (2001) Biochemistry

40, 11552-11558

17. Davydov, R. M., Yoshida, T., Ikeda-Saito, M., and Hoffman, B. M. (1999) J. Am. Chem.

Soc. 121, 10656-10657

18. Sono, M., Roach, M. P., Coulter, E. D., and Dawson, J. H. (1996) Chem. Rev. 96, 2841-

2887

19. Mansfield Matera, K., Takahashi, S., Fujii, H., Zhou, H., Ishikawa, K., Yoshimura, T.,

Rousseau, D. L., Yoshida, T., and Ikeda-Saito, M. (1996) J. Biol. Chem. 271, 6618-6224

20. Takahashi, S., Ishikawa, K., Takeuchi, N., Yoshida, T., Ikeda-Saito, M., and Rousseau, D. L.

(1995) J. Am. Chem. Soc. 117, 6002-6006

21. Migita, C. T., Mansfield Matera, K., Ikeda-Saito, M., Olson, J. S., Fujii, H., Yoshimura, T.,

Zhou, H., and Yoshida, T. (1998) J. Biol. Chem. 273, 945-949

22. Fujii, H., Dou, Y., Zhou, H., Yoshida, T., and Ikeda-Saito (1998) J. Am. Chem. Soc. 120,

8251-8252

23. Sugishima, M., Sakamoto, H., Kakuta, Y., Omata, Y., Hayashi, S., Noguchi, M., and

Fukuyama, K. (2002) Biochemistry 41, 7293-7300

24. Schuller, D. J., Wilks, A., Ortiz de Montellano, P. R., and Poulos, T. L. (1999) Nature

Struct. Biol. 6, 860-867

25. Davydov, R. M., Kofman, V., Fujii, H., Yoshida, T., Ikeda-Saito, M., and Hoffman, B. M.

(2002) J. Am. Chem. Soc. 124, 1798-1808

26. Lad, L., Wang, J., Li, H., Friedman, J., Bhaskar, B., Ortiz de Montellano, P. R., and Poulos,

T. L. (2003) J. Mol. Biol. 330, 527-538

27. Rohlfs, R. J., Mathews, A. J., Carver, T. E., Olson, J. S., Springer, B. A., Egeberg, K. D.,


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

18

and Sligar, S. G. (1990) J. Biol. Chem. 265, 3168-3176

28. Takahashi, S., Yeh, S.-R., Das, T. K., Chan, C.-K., Gottfried, D. S. & Rousseau, D. L.

(1997) Nature Struct. Biol. 4, 44-50

29. Couture, M., Stuehr, D. J., and Rousseau, D. L. (2000) J. Biol. Chem. 275, 3201-3205

30. Otwinowski, Z., and Minor, W. (1997) Methods Enzymol. 276, 307-326

31. Brunger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P., Grosse-Kunstleve, R.

W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Reak, R. J., Rice, L. M., Simonson,

T., and Warren, G. L. (1998) Acta Crystallogr. D54, 905-921

32. CCP4: Collaborative Computational Project, Number 4 (1994) Acta Crystallogr. D50, 760-

767

33. Jones, T. A., Zou, J.-Y., Cowan, S. E., and Kjeldgaard, M. (1991) Acta Crystallogr. A 47,

100-119

34. Kissinger, C. R., Gehlhaar, D. K., and Fogel, D. B. (1999) Acta Crystallogr. D55, 484-491

35. Schomaker, V. and Trueblood, K.N. (1968) Acta Cryst., B24, 63-76

36. Kraulis, P. J. (1991) J. Appl. Crystallogr. 24, 946-950

37. Esnouf, R. M. (1999) Acta Crystallogr. D55, 938-940

38. Springer, B. A., Sligar, S. G., Olson, J. S., and Phillips, G. N., Jr. (1994) Chem. Rev. 94,

699-714

39. Krzywda, S., Mushudov, G. N., Brzozowski, A. M., Jaskowlski, M., Scott, E. E., Klizas, S.

A., Gibson, Q. H., Olson, J. S., and Wilkinson, A. J. (1998) Biochemistry 37, 15896-15907

40. Gibson, Q. H., and Smith, M. H. (1965) Proc. R. Soc. Lond. B Biol. Sci. 163, 206–214

41. Draghi, F., Miele, A. E., Travaglini-Allocatelli, C., Vallone, B., Brunori, M., Gibson, Q. H.,

and Olson, J.S. (2002) J. Biol. Chem. 277, 7509-7519

42. Brucker, E. A., Olson, J. S., Phillips, G. N., Jr, Dou, Y., and Ikeda-Saito, M. (1996) J. Biol.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

19

Chem. 271, 25419-25422

43. Denisov, I. G., Ikeda-Saito, M., Yoshida, T., and Sligar, S. G. (2002) FEBS Lett. 537, 203-

206

44. Vojtêchovský, J., Chu, K., Berendzen, J., Sweet, R. M., and Schlichting, I. (1999) Biophys.

J. 77, 2153-2174

45. Condon, P. J., and Royer Jr., W. E. (1994) J. Biol. Chem. 269, 25259 (1994)

46. Schlichting, I., Berendzen, J., Chu, K., Stock, A. M., Maves, S. A., Benson, D. E., Sweet, R.

M., Ringe, D., Petsko, G. A., Sligar S. G. (2000) Science 287, 1615-1622

47. Berglund, G. I., Carlsson, G. H., Smith, A. T. Szöke, H., Henriksen, A., and Hajdu, J.

(2002) Nature 417, 463-468

48. Fujii, H., Zhang, H., Tomita, T., Ikeda-Saito, M., and Yoshida, T. (2001) J. Am. Chem. Soc.

123, 6475-6484

49. Davydov, D. Matsui, T., Fujii, H., Ikeda-Saito, M., and Hoffman, B. M. (2003) J. Am.

Chem. Soc. 125, 16208-16029

50. Zhang, X., Fujii, H., Mansfield Matera, K., Migita, C. T., Sun, D., Sato, M., Ikeda-Saito, M.,

and Yoshida, T. (2003) Biochemistry 42, 7418-7426

51. Harutyunyan, E. H., Safonova, T. N., Kuranova, I. P., Popov, A. N., Teplyakov, A. V.,

Obmolova, G. V., Rusakov, A. A., Vainshtein, B. K., Dodson, G. G., Wilson, J. C., and

Perutz, M. F. (1995) J. Mol. Biol. 251, 104–115

52. Hargrove, M. S., Barry, J. K., Brucker, E. A., Berry, M. B., Phillips, G. N., Jr., Olson, J. S.,

Arredondo-Peter, R., Dean, J. M., Klucas, R. V., and Sarath, G. (1997) J. Mol. Biol. 266,

1032–1042

53. Yang, J., Kloek, A. P., Goldberg, D. E., and Mathews, F. S. (1995) Proc. Natl. Acad. Sci. U.

S. A. 92, 4224–4228


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

20

54. Scheidt, W. R., and Chapman, D. M. (1986) J. Am. Chem. Soc. 108, 1163-1167

55. Phillips, G. N. Jr., Teodoro, M. L., Li, T., Smith, B., and Olson, J. S. (1999) J. Phys. Chem.

B 103, 8817-8829

56. Smerdon, S.J., Dodson, G.G., Wilkinson, A.J., Gibson, Q.H., Blackmore, R.S., Carver, T.E.,

and Olson, J.S., (1991) Biochemistry 30, 6252-6260

57. De Baere, I., Perutz, M. F., Klinger, L., Marden, M. C., and Poyart, C. (1994) Proc. Natl.

Acad. Sci. U. S. A. 91, 1594-1597

58. Olson, J. S. and Phillips, G. N., Jr. (1997) J. Biol. Inorg. Chem. 2, 544-552


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

21

Table I. Statistics of data collection and structure refinement

__________________________________________________ Data Collection Resolution range (Å) 50.0-1.85 (1.92-1.85) Total no. of reflections 231185 No. of unique reflections 60627 Mosaicity (degree) 0.18 I/σ 22.0 (4.1) Completeness (%) 100.0 (100.0) Rmerge

a 0.071 (0.343) Structure refinement Resolution range (Å) 40.0 - 1.85 No. of non-hydrogen protein atoms 5686 No. of water molecules 449 R-factorb 0.153 Rfree

b 0.192 Mean B-values (Å2) All atoms 13.85 Proteins 11.98 Hemes 10.44 Waters 28.67 rms deviations from ideal geometry Bond lengths (Å) 0.020 Bond angles (degree) 1.785 __________________________________________________ aRmerge = Σ |I - <I>|/ΣI

bR = Σ (|Fobs| - |Fcalc|)/Σ|Fobs|. The Rfree is the R calculated on the 10% reflections excluded for

refinement.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

22

Table II. Parameters for O2 binding to the HmuO and rat HO-1 heme complexes, pig Mb, and

Ascaris Hb in 0.1 M phosphate, pH 7, 20 o C. The rate constants for rat HO-1 are taken from

Migita et al. (21), pig wild-type and V68N Mb from Kryzwda et al. (38) and that for Ascaris Hb

from Gibson and Smith (39), and Draghi et al. (42). Errors for HO-1 were obtained from 3

different, independent experimental determinations (different samples and different days) and

indicate that there are no significant differences between the ligand binding parameters for

HmuO and HO-1. The errors in the myoglobin rate constants are ~ ±20% as has been observed

for sperm whale myoglobin (38).

Protein O2 Binding CO Binding Ratio

k'O2

(µM-1 s-1)

kO2

(s-1)

KO2

(µM-1)

k'CO

(µM-1 s-1)

kCO

(s-1)

KCO

(µM-1) KCO/KO2

HmuO 4.7 0.22 21 ~ 0.6a 0.004 150 7.1

HO-1 6.9±1.5 0.25±0.10 28±13 ~ 0.9±0.2a 0.009 100±25 3.6±1.9

pig Mb 17±3 14±3 1.2±0.3 0.78±0.16 0.019±0.003 41±10 34±12

pig V68N Mb 3.0 0.67 4.5 0.12 0.011 11 2.4

Ascaris Hb 1.5 0.004 370 0.21 0.018 12 0.03

aThe time courses for CO rebinding to HmuO and HO-1 were moderately biphasic and could be

fitted to a two exponential expression with roughly equal amplitudes. The fast and slow phase

bimolecular rate constants for CO binding were 0.96 µM-1 s-1 and 0.29 µM-1 s-1 for HmuO and

1.3 µM-1 s-1 and 0.64 µM-1 s-1 for HO-1. The values reported in the table were obtained by

fitting the time courses to a single exponential expression at several different CO concentration.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

23

Figure Legends

Figure 1. Schematics of heme oxygenase catalytic intermediates.

Figure 2. Resonance Raman spectra of the oxygen-bound form of the heme complex of

HmuO. A, the 16O2-bound HmuO complex; B, 18O2-bound HmuO complex; C, Difference

spectrum between A and B; D. 16O2 - 18O2 difference spectrum of the oxy form of the heme

complex of rat HO-1 (20).

Figure 3. Structure of the oxygen binding site. The final σA-weighted 2Fo-Fc map (blue) at

1.8 σ level and the simulated annealing omit Fo-Fc map (red) at 4.2 σ level.

Figure 4. Structure of the immediate vicinity of the bound dioxygen. The distal pocket Gly

residues and the bound dioxygen are shown in CPK model, and the heme is illustrated in a ball

and stick diagram. Carbon, nitrogen, and oxygen atoms of the heme and Gly residues are shown

in gray, blue, and red, respectively. The bound dioxygen molecule is shown in magenta. Top,

view from the pyrrole A; bottom, view from the distal side of the heme.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

351

541

415 565

404343

404344 539

566414351

347

563

413

543410

677

A

300 400 500 600 700Raman Shift (cm -1)

B

C

D365

310

299

299

310

365

��


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Q46Q46D136 D136

G139 G139

W1 W1

W2 W2W3W3

W4W4

Figure 3


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

α-meso

α-meso

Gly135Gly139

Gly140

Gly139

Gly135

Figure 4


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Supplemental data. Optical absorption spectra of oxy HmuOcrystals before (top) and after (bottom) the X-ray diffraction mea-surements. Spectral shapes are slightly different, because thebaselines are not matched due to the difference in alignment of thecrystals with respect to the incident measuring light beam and to theformation of the ice on the surface of the crystals.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Rousseau, John S. Olson and Masao Ikeda-SaitoMasaki Unno, Toshitaka Matsui, Grace C. Chu, Manon Couture, Tadashi Yoshida, Denis L.

diphtheriae: Implications for heme oxygenase functionCrystal structure of the dioxygen-bound heme oxygenase from Corynebacterium

published online February 13, 2004J. Biol. Chem.

10.1074/jbc.M400491200Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

http://www.jbc.org/content/suppl/2004/02/17/M400491200.DC1


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M400491200

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;M400491200v1&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2004/02/13/jbc.M400491200.citation

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=early/2004/02/13/jbc&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2004/02/13/jbc.M400491200.citation

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2004/02/17/M400491200.DC1

http://www.jbc.org/

Documents

Crystal Structure of the Dioxygen-bound Heme Oxygenase from