1. Journal of Inorganic Biochemistry 91 (2002) 635–643
www.elsevier.com/locate/jinorgbio
Mesopone cytochrome c peroxidase: functional model of heme oxygenated
oxidases
a ,1 b ,1 a b a ,
*Chad E. Immoos , B. Bhaskar , Michael S. Cohen , Tiffany P. Barrows , Patrick J. Farmer ,
b ,
*T.L. Poulos
a
Department of Chemistry, University of California, Irvine, CA 92697-2025, USA
b
Department of Molecular Biology and Biochemistry, Physiology and Biophysics, and the Program in Macromolecular Structure,
University of California, Irvine, CA 92697-3900, USA
Received 20 December 2001; received in revised form 13 February 2002; accepted 15 April 2002
Abstract
The effect of heme ring oxygenation on enzyme structure and function has been examined in a reconstituted cytochrome c peroxidase.
Oxochlorin derivatives were formed by OsO treatment of mesoporphyrin followed by acid-catalyzed pinacol rearrangement. The4
1
northern oxochlorin isomers were isolated by chromatography, and the regio-isomers assignments determined by 2D COSY and NOE H
NMR. The major isomer, 4-mesoporphyrinone (Mp), was metallated with FeCl and reconstituted into cytochrome c peroxidase (CcP)2
forming a hybrid green protein, MpCcP. The heme-altered enzyme has 99% wild-type peroxidase activity with cytochrome c. EPR
191
spectroscopy of MpCcP intermediate compound I verifies the formation of the Trp radical similar to wild-type CcP in the reaction
cycle. Peroxidase activity with small molecules is varied: guaiacol turnover increases approximately five-fold while that with ferrocyanide
III II
is |85% of native. The electron-withdrawing oxo-substitutents on the cofactor cause a |60-mV increase in Fe /Fe reduction potential.
The present investigation represents the first structural characterization of an oxochlorin protein with X-ray intensity data collected to 1.70
˚A. Although a mixture of R- and S-mesopone isomers of the FeMP cofactor was used during heme incorporation into the apo-protein,
only the S-isomer is found in the crystallized protein.
 2002 Elsevier Science Inc. All rights reserved.
Keywords: Heme cd model; Reconstituted oxochlorin; Cytochrome c peroxidase1
1. Introduction rings is found in cytochrome heme d, the terminal oxidase
of Escherichia coli, which reduces oxygen to water using
The heme cd enzymes are unusual bifunctional ubiquinol as an electron donor [3,4]. Synthetic models1
catalysts for both the single-electron reduction of nitrite to have been studied to determine the effect of heme oxygen-
nitric oxide and the four-electron reduction of oxygen to ation on structural, spectroscopic and electrochemical
water [1]. The oxidase activity is especially unusual, and is properties [5–8]. The nitrite reductase reactivity of oxoch-
induced under conditions of low oxygen concentrations. lorin model complexes has been evaluated [9,10], but little
The oxygen-binding site has been identified as the heme d has been done to examine the effect of heme modification1
cofactor, an unusual Fe-dioxoisobacteriochlorin cofactor on oxidase activity.
(Scheme 1) [2]. Another d-type heme with oxygenated The catalytic oxidase sequence is thought to go through
IV 1
a ferryl porphyrin radical intermediate, Fe =O(R ), in
analogy to a commonly accepted scheme for cytochrome
oxidase. There is some indication that heme oxygenation
*Corresponding authors. Tel.: 11-949-824-6079; fax: 11-949-824- may affect these reactivities. For instance, the ferrous
2210. oxy-adduct is unusually stable in cytochrome bd oxidase
*Corresponding authors. Tel.: 11-949-824-7020; fax: 11-949-824-
[11], and a long-lived intermediate has been seen during
3280. IV
turnover, suggested to be a stable ferryl, Fe =O [12]. AsE-mail addresses: pfarmer@uci.edu (P.J. Farmer), poulos@uci.edu
yet, there is little consensus as to why such heme oxygena-(T.L. Poulos).
1
Equal contributions. tions have evolved for these catalytic functions.
0162-0134/02/$ – see front matter  2002 Elsevier Science Inc. All rights reserved.
PII: S0162-0134(02)00447-6

2. 636 C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643
Scheme 1.
II 2 1 III 2
Fe 1 O 1 e 1 H → Fe 2 HO (1) 2. Experimental procedures2 2
III 2 1 IV 1
Fe 2 HO 1 H → Fe =O(R ) 1 H O (2) All solvents were of ACS chemical grade (Fisher) and2 2
were used without further purification unless otherwise
IV 1 2 1 III
Fe =O(R ) 1 2e 1 2H → Fe 1 H O (3) indicated. Reagent grade methylene chloride was dried by2
distillation from calcium hydride. Dimethylformamide and
III IV 1
Fe 1 H O → Fe =O(R ) 1 H O (4) pyridine were distilled from CaO. Mesoporphyrin dimethyl2 2 2
ester was purchased from Porphyrin Products. Osmium
Peroxidases also go through ferryl states during catalytic tetroxide and dimethyldidodecylammonium bromide
III
turnover, via reaction of the ferric, Fe , state with (ddab) were purchased from Acros. Water was purified to a
21
hydrogen peroxide [Eq. (4)] [13]. The second oxidation specific resistance of 18 mV in a Barnstead nanopure
1
equivalent typically forms a porphyrin radical (R ), but in water purification system.
the case of cytochrome c peroxidase, a stable protein Enzymes and reagents were purchased from Roche
191
radical is formed on Trp [14]. The two oxidizing Molecular Biochemicals and New England Biolabs (Bev-
1 1
equivalents thus stored in the enzyme are used to carry out erly, MA). Chromatography columns, media and Na /K
two one-electron oxidations of various substrates which, in free Tris base were purchased from Amersham-Pharmacia
the case of cytochrome c peroxidase (CcP), is ferrocytoch- Biotech. 2-Methyl-2,4-pentanediol (MPD) and ion free
rome c. phosphoric acid were purchased from Aldrich Chemicals.
Previous comparison of the redox properties of Fe- Horse heart and yeast cytochrome c and guaiacol were
complexes of oxochlorins and dioxoisobacteriochlorins purchased from Sigma. All other chemicals were molecu-
showed that the oxochlorins undergo macrocycle oxida- lar biology grade or better and were purchased from Sigma
tions at nearly the same potential as the parent porphyrin or Fisher.
III II 1 13
complexes [5]. However, metal-centered Fe -Fe reduc- H NMR and C NMR spectra were measured on a
tions become easier upon oxygenation of the ring, agreeing Bruker 500 spectrometer at 500 MHz. Chemical shifts
with Hueckel calculations showing an increased charge on were referenced to CDCl at 7.24 and 77.00 ppm. Mass3
the metal as oxo groups are added to the porphyrin spectra were taken on an electro-spray (ES) mass spec-
skeleton [7]. Introduction of oxo functions might thus trometer (Micromass LCT). A Hewlett-Packard 8453
provide a mode of controlling the reactivity of the high UV–Vis spectrophotometer was used for all optical mea-
oxidation state iron intermediates, such as the ferryl, surements.
during enzymatic activity [15].
For a better understanding of the effect of heme ring 2.1. Synthesis of dihydroxymesochlorin dimethyl ester
oxygenation on enzymatic function, we have synthesized (mesochlorindiol)
an oxochlorin derivative, Fe-mesopone (FeMp) and recon-
stituted it into apo-CcP, forming the hybrid protein MpCcP. Mesoporphyrin dimethyl ester (487 mg, 0.82 mmol) was
The altered cofactor in the hybrid results in a selective dissolved in degassed CH Cl (200 ml). A few drops of2 2
enhancement of its peroxidase activity with certain sub- ether were added, and osmium tetroxide (250 mg, 0.98
strates. The results of a single crystal structural determi- mmol) was quickly added to the solution. Dry pyridine
nation offer the first structural parameters for a reconsti- (0.6 ml) was subsequently added and the mixture was
tuted oxochlorin protein, and confirm that the altered allowed to stir at room temperature, under a nitrogen
activity is not due to large structural changes in the atmosphere, in the dark for 20 h. The solvents were then
reconstituted hybrid. removed by rotary evaporation and the dark-green residue

3. C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643 637
was dissolved in CH Cl (30 ml) and MeOH (100 ml). the major fraction (45.7 mg, 80%). UV–Vis l 396 nm,2 2 max
Hydrogen sulfide, generated in situ from acid and sodium 550 (broad), 710.
sulfide, was passed through the solution for 20 min in
order to decompose the osmium ester. The solvents were 2.4. Protein expression, reconstitution and purification
removed by rotary evaporation and the crude product was
dissolved in CH Cl and filtered to remove the precipitated Wild-type CcP (WTCcP) was expressed using a T72 2
osmium sulfide. The crude diol was separated on a silica promoter in Escherichia coli BL21(DE3) cells induced at
gel column. Unreacted mesoporphyrin was eluted first with A of 1.2–1.5 with 750 mM IPTG. Proteins were purified600
1% THF in CH Cl (100 mg, 21%). The green mesoch- as previously described by Fishel et al. [16] and2 2
lorindiol mixture (four different regio-isomers) was eluted Choudhury et al. [17] using similar gradient parameters.
next with 3% MeOH in CH Cl . The mesochlorindiol After gel filtration on a Sephadex G-75 column, apo-CcP2 2
isomeric mixture was collected and concentrated in vacuo fraction was separated from small amounts of holo-CcP by
without further purification (300 mg, 60%). UV–Vis l HiTrap-Q anion-exchange chromatography on Amersham-max
391 nm, 495, 523, 590, 642. ESIMS, m/z 629.4 (M1Na, Pharmacia FPLC system. Stepping the gradient first to 80
651.38, 100). mM, then to 130 mM potassium phosphate, pH 6.0 (KPB),
and finally a linear gradient from 130 to 500 mM KPB, pH
6.0 separated apo-CcP from holo-CcP. Apo-CcP dialyzed
2.2. Synthesis of 4-mesoporphyrinone dimethyl ester III
extensively against 100 mM KPB, pH 7.2 before Fe -
(mesopone or 8,12,-diethyl-3,8,13,17-tetramethyl-7-oxo- III
mesopone incorporation. After dissolving Fe -mesopone
prophyrin-2,18-diproprionic acid dimethyl ester)
chloride in a mixture of 0.1 N NaOH–methanol, it was
diluted by the addition of 100 mM KPB, pH 7.2. This
Mesochlorindiol (300 mg, 0.48 mmol) was dissolved in
mixture was added to the dialyzed apo-CcP, followed by
dry CH Cl (150 ml). A few drops of H SO were added2 2 2 4
incubation in the cold for 12 h and extensive dialysis
to the CH Cl solution and the mixture was allowed to stir2 2
against 20 mM KPB, pH 6.0. MpCcP was further purified
at room temperature under nitrogen for 20 min upon which
on DEAE-Sephacel anion-exchange column. MpCcP was
the solution turned dark-blue. The reaction mixture was
stored in 200 mM KPB, pH 6.0 at 4 8C after extensive
poured into an ice bath and washed with brine solution
dialysis against the same buffer. CcP concentrations were
(33200 ml) and 10 mM sodium bicarbonate (23200 ml).
estimated spectrophotometrically using an extinction co-
The organic layer was dried (Na SO ) and concentrated in 21 212 4
efficient at 408 nm (´ ) of 96 mM cm .408
vacuo to obtain a purple solid. Column chromatography
(silica gel, eluent 1% THF in CH Cl ) was used to isolate2 2
2.5. Crystallization, X-ray data collection and structure
the desired mesoporphyrinone isomer, 4-mesopor-
refinement1
phyrinone (100 mg, 35%). H NMR (CDCl ) d 0.43 (3 H,3
t, sat. ethyl), 1.84 (3 H, t, ethyl), 2.08 (3 H, s, Me sat.),
Diffraction quality crystals were prepared in 30% 2-
2.75 (2 H, q, sat. ethyl), 3.21 (4 H, m, –CH CH OOCH ),2 2 3
methyl-2,4-pentanediol (MPD), 50 mM KPO , pH 6.04
3.44 (3 H, s, Me), 3.56 (3 H, s, ring-Me), 3.58 (3 H, s,
according to Edwards and Poulos [18] as later modified by
Me), 3.65 (3 H, s, CO CH ), 3.68 (3 H, s CO CH ), 3.962 3 2 3
Sundaramoorthy et al. [19] An initial concentration of 500
(2 H, q, ethyl), 4.18 (2 H, t, –CH CH OOCH ), 4.35 (2 212 2 3
mM (|16 mg ml ) MpCcP was used to grow smaller 0.2
H, t, –CH CH OOCH ), 9.12 (1 H, s, a-meso), 9.79 (1 H,2 2 3
mm crystals from touch seeding. This smaller size crystal
s, d-meso), 9.84 (1 H, s, b-meso), 9.90 (1 H, s, g-meso).
was necessary for proper cryogenic freezing and to avoid13
C NMR (CDCl ) d 211.07 (carbonyl). UV–Vis l3 max
twinning during crystal growth. Crystals belong to the
(´ ) 404 nm (137), 508 (7.2), 545 (9.2), 585 (4.8), 641mM
same space group as that of WTCcP, P2 2 2 with a51 1 1
(27.1). ESIMS, m/z 611.39 (M1H, 100). ˚107.011, b576.091, c551.124 A, with one CcP molecule
per asymmetric unit. X-ray intensity data to a Bragg
III ˚2.3. Synthesis of Fe -Cl mesopone spacing of 1.70 A were collected from one flash frozen
crystal using an R-AXIS IV imaging plate and a Rigaku
4-Mesoporphyrinone dimethyl ester was hydrolyzed by rotating anode X-ray source equipped with a Crystal Logic
dissolving in 2 N NaOH:THF and stirring for 24 h in the cryogenic N delivery system. Initial image processing,2
dark. The iron complex was generated by refluxing 4- indexing, and integration were performed with DENZO
mesoporphyrinone diacid (49.3 mg, 0.085 mmol) and (version 1.9.1), and the integrated data were scaled using
FeCl ?4H O (33.4 mg, 0.17 mmol) in dry DMF under SCALEPACK (version 1.9.0.) [20]. A summary of data2 2
nitrogen for 6 h. The solvent was removed by vacuum collection is provided in Table 1.
distillation and the solid was washed with slightly acidic The starting model for refinement was WTCcP refined
water and filtered to remove excess FeCl ?4H O. The using ultra-high resolution data obtained at cryogenic2 2
resulting black solid was dissolved in MeOH and purified temperatures at Stanford Synchotron Radiation Laboratory
on a Sephadex LH-20. The final product was collected as (SSRL). The model was fitted to MpCcP data using rigid

4. 638 C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643
21
Table 1 cm . Ferrocyanide peroxidase activity was determined in
Statistics of data collection and refinement of MpCcP (PDB code: 1KOK)
17 mM K Fe(CN) ?3H O, 600 mM H O and 10 nM4 6 2 2 2
Data collection enzyme in 100 mM Tris–phosphate, pH 6.0 [27]. The
Space group P2 2 21 1 1 formation of oxidized product, ferricyanide, was monitored
˚Unit cell dimension A a5107.011, b576.091, c551.124 21
at 420 nm using an extinction coefficient ´ of 1000 M420Total reflections 423,704 21
cm .Unique reflections 46,346
˚Resolution (A) 1.70
I/s last shell 4.97 2.7. Voltammetry
Completeness full dataset (%) 99.1
Completeness last shell (%) 90.1
A BAS 100-W electrochemical analyzer was used with aa
R-factor (R ) 0.046symm
standard three-electrode cell, containing a Pt wire counter
electrode, an Ag/AgCl reference electrode (BAS), and aRefinement
˚Resolution range (A) 50.0–1.70 basal-plane PG disk working electrode. All electrochemi-
Number of reflections 45,673 cal experiments were performed at room temperatureb
R 0.1859cryst
(2262 8C). The electrodes were immersed into 50 mM
R (5% data set aside) 0.2072free
c potassium phosphate buffer solutions containing 0.1 M˚r.m.s bond lengths (A) 0.0076
c
NaBr electrolyte. In order to remove oxygen prior tor.m.s bond angles (8) 1.2576
beginning the electrochemical experiments, the analyte
Number of atoms solution was purged with purified nitrogen for 20 min. A
Protein 2374
nitrogen atmosphere was maintained over the solution
Prosthetic group 43
during the experiments. A 10-ml portion of a 0.01-M ddabWater 634
III
solution (in water) and 10 ml of a 0.5-mM MpCcP (Fe )a
R 5o uI 2I u/o I .sym h h h
b solution were cast onto homemade electrodes made ofR 5o (uF u2uF u)/o uF u.cryst obs calc obs
c basal-plane, pyrolytic graphite (PG) discs (Union Carbide)R.m.s. bond and r.m.s. angle represent the root-mean-squared devia-
tion between the observed and ideal values. sealed with epoxy into glass tubes. The MpCcP/ddab films
were dried in a closed vessel overnight and then exposed
to air for 24 h.
body refinement as implemented in the CNS (version 1.0)
[21]. Next, 5% of the data were set aside to calculate R 2.8. Electron paramagnetic resonance (EPR)free
for cross-validation [22]. To begin refinement of the spectroscopy
model, simulated annealing starting at 3000 K as im-
plemented in CNS was used to remove any previous model EPR spectra were recorded on a Bruker ESP300 spec-
bias. The subsequent refinements involved multiple rounds trophotometer equipped with an Air Products LTR3 liquid
of model building in TOM (version 3.0) [23], and then helium cryostat. Experimental conditions used to record
191
positional refinement for 100 cycles followed by B-factor Trp cation radical formation of the wild-type protein
refinement for 30 cycles with CNS. In addition 634 and mutants were as follows: microwave frequency, 9.475
ordered solvent molecules were added to the molecular GHz; microwave power, 0.5 mW; modulation amplitude,
model. The final refinement parameters of MpCcP are 4.57 G; modulation frequency, 100 kHz; field sweep rate,
21
summarized in Table 1. 11.92 G s ; time constant, 0.0256 ms; and receiver gain,
4
1.0310 . Compound I was formed by the addition of an
2.6. Steady state activity assays equal volume of 360 mM H O in respective buffers and2 2
the samples were frozen immediately in quartz EPR tubes
The steady state oxidation of horse heart and yeast by submersion in a mixture of n-hexanes chilled into a
cytochrome c (ferrocytochrome c) was measured at room slurry with liquid nitrogen over a period of 40 s. Spectra
temperature in a Cary 3E UV-visible spectrophotometer were recorded at 4 K. The data obtained were an average
21 21
using a D´ of 19,600 M cm . Typical final reactions of ten scans.550
consisted of 25–30 mM dithionite-reduced horse heart or
yeast cytochrome c, 180 mM H O and 250 pM enzyme in2 2
100 mM Tris–phosphate, pH 6.0 [24]. Hydrogen peroxide 3. Results and discussion
concentrations were standardized with KMnO using the4
method of Fowler and Bright [25]. A modified route was employed to synthesize 4-
The guaiacol peroxidase activity was determined in 100 mesoporphyrinone, based on the work of Chang and Smith
mM Tris–phosphate, pH 6.0 containing 100 mM guaiacol, (Scheme 2) [28–30]. Treatment of mesoporphyrin di-
600 mM H O and 10 nM enzyme [26]. The formation of methyl ester with osmium tetraoxide yielded mixtures of2 2
the oxidized product, tetraguaiacol, was followed at 470 isomers with dihydroxyl substituents at all the pyrrole
21
nm using an extinction coefficient ´ of 26,600 M positions, acidification induced a pinacol rearrangement,470

5. C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643 639
Scheme 2.
IV
resulting in the corresponding mixture of oxochlorin peroxide yielded compound I, Fe =O, a form that has a
products. The northern oxochlorin isomers were isolable Soret absorbance similar to the WTCcP, with an absor-
by chromatography, and assignments of specific oxoch- bance |600 nm typical of chlorins.
lorin regio-isomers were determined by 2D COSY and Activity assays with the native substrate, cytochrome c,
NOE experiments, using the distinctive NOE patterns of show that the cofactor-modified enzyme MpCcP has full
the beta-methylene protons, as illustrated in Scheme 3. IR wild-type activity. As the native activity is mediated by a
191
spectra of both northern products exhibits a new carbonyl stable protein-based radical on Trp in the compound I
21
stretch at 1715 cm indicative of ketone formation. After state, which forms part of the cytochrome c binding
purification of the major isomer, de-esterification and proximal loop of CcP, this result suggests both the rate of
metallation of the free base oxochlorin yielded the formation and the stability of the Trp cation radical of the
Fe(Mp)Cl in |15% overall yield from the diol. cofactor-modified and native enzymes are similar.
Titration of apo-CcP solutions with aliquots of To better characterize the compound I state, EPR spectra
Fe(Mp)Cl results in the growth of a sharp Soret-band in of the MpCcP and WTCcP were compared under the same
UV-visible spectrum of the hybrid adduct, MpCcP, which experimental conditions, as seen in Fig. 2. The signal of
191
is close to the WTCcP absorbance at 408 nm. Typically, an compound I in native CcP is due to a Trp radical, which
IV
approximately four-fold excess of cofactor was needed for is in magnetic exchange with the S52 Fe =O of the ferryl
complete reconstitution. The mixture was purified by porphyrin. Under our conditions, the WTCcP shows an
standard methods, and the purified protein characterized by apparent axial signal, with g-values of g 52.0374 anduu
UV–visible absorption spectroscopy in its relevant oxida- g 52.0061; the MpCcP spectra display a very similar, but'
II
tion states (Table 2). The green ferrous form, Fe MpCcP, slightly shifted signal, with g-values of g 52.00330 anduu
has a quasi-split Soret at 410 and 432 nm and a sharp peak g 52.0055. The EPR signal of compound I state of'
191
at 607 nm that is typical of chlorin hemes (Fig. 1). MpCcP is indicative of a Trp radical quite similar to
Treatment with approximately stoichiometric amounts of that of the wild-type enzyme, in good agreement with the
maintenance of native activity with cytochrome c.
The ability of MpCcP to oxidize small molecule sub-
Scheme 3.
Table 2
Comparison of absorption peak maxima (nm)
WTCcP MpCcP
31
Ferric (Fe ) 408, 505, 645 408, 560, 597
21
Ferrous (Fe ) 439, 559, 588(sh) 410(sh), 432, 607
Fig. 1. Absorbance spectra of 5 mM ferric ( ), ferrous ( ), compound I41
Compound I (Fe ) 420, 530, 561 420, 565(sh), 601
(-----) MpCcP in 50 mM phosphate buffer, pH 6.0.

6. 640 C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643
Fig. 2. EPR spectra of compound I oxidation states of 300 mM WTCcP
]]]] ]]]](----) and MpCcP ( ), in 50 mM Tris–phosphate, pH 6.0 formed by the Fig. 3. Cyclic voltammograms of WTCcP ( ) and MpCcP (----) in
21
addition of 360 mM H O . Instrumental conditions are described in the ddab-film electrodes (100 mV s , 100 mM NaBr, 50 mM phosphate2 2
text. Dotted lines show the small shift in apparent g-parallel component. buffer, pH 7.0).
strates was determined using guaiacol and potassium ferricyanide is most likely due to differences in substrate
ferrocyanide as substrates. As shown in Table 3, the binding sites. (A reviewer has suggested that the low
activity with guaiacol increases approximately five-fold, turnover number for ferricyanide may be due to a differ-
while that of ferrocyanide is |85% of that of WTCcP. The ence in the competitive inhibition of the catalysis by
increased activity with guaiacol may be due to a change in cyanide in the native and hybrid enzyme, which is known
IV
oxidizing potential of the Fe =O state of the modified to be generated during the assay.) Ortiz de Montellano has
cofactor. To compare the effect of heme modification on demonstrated that small molecule substrates are oxidized at
the redox potentials of the heme sites, we utilized the sites close to the exposed d-meso heme edge [35]. Recent
surfactant film methodology of Rusling [31–33]. The work has shown that mutations on the protein surface near
enzymes were cast in thin films of ddab on basal plane the heme edge can have a substantial effect on small
graphite, and the electrochemical response determined in molecule turnover in CcP [36]. Therefore, we conclude
IV
aqueous solutions (Fig. 3). Although the Fe =O state that increased activity of MpCcP toward guaiacol is due to
III II
could not be assessed by this method, both Fe /Fe and the higher redox potential and/or lower kinetic barrier to
II I IV
Fe /Fe couples in MpCcP are shifted to more positive reduction of the Fe =O center in the modified heme.
potentials relative to WTCcP (Table 4). The increase in However, since large substrates like ferrocytochrome c and
Fe-based potential is as expected for a cofactor with ferrocyanide interact near the electron entry point to the
191
electron-withdrawing substituents, and likewise would Trp cationic radical and not at the heme edge, reactivity
suggest a higher activity if the electron transfer is rate- toward these substrates is much closer to that of WTCcP.
limiting. In contrast, the reduction potential for an oxoch- To assess the structural changes caused by the heme
lorin-substituted cytochrome b5 was unchanged from that alteration, diffraction quality crystals of MpCcP were
of the native system, but like MpCcP, the heme-altered grown using standard methods, and full structural analysis
cytochrome b5 was capable of transferring electrons to was performed. An omit F 2 F difference map of theo c
cytochrome c in a reconstituted system simulating the active site of MpCcP is shown in Fig. 4. The map shows
native activity [34]. the characteristic distal and proximal heme domains with
175 191
The variable reactivity of MpCcP with guaiacol and His ligating the modified heme and Trp , which forms
the cationic radical in the reaction cycle. The map also
Table 3 shows the modified heme with the ethyl group in the
Comparison of catalytic activity
pyrrole ring between a- and b-meso edges of the protopor-
21
Substrate k (s )cat
WTCcP MpCcP
Table 4
a
HH cytochrome c 875.85 871.54 Comparison of reduction potentials derived from direct electrochemical
b
Guaiacol 0.953 5.18 measurements in ddab films for MpCcP and WTCcP (mV vs. NHE)
a
K Fe(CN) 160.34 140.2 III / II II / I4 6
Fe Fe
a
Horse heart cytochrome c and ferricyanide assay in 100 mM Tris–
WTCcP 119 2779
phosphate, pH 6.0.
MpCcP 181 2733b
Guaiacol assay in 50 mM Tris–phosphate, pH 6.0.

7. C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643 641
Fig. 4. Active site structure of MpCcP, showing saddling of the heme and its orientation with respect to the distal pocket residues.
phyrin instead of the vinyl group found in protoporphyrin side chains are fully extended, they are prevented from
IX. The reconstituted mesopone occupies a position similar extensive contact with the surrounding medium by a rigid
to the heme in WTCcP. Although a mixture of R- and network of hydrogen bonds.
S-mesopone isomers was used during heme incorporation The modified heme, like the regular heme of WTCcP, is
into the apo-protein, only the S-isomer is found in the not planar but slightly distorted into a saddle shape, with
crystallized protein. The reason that only the S-isomer pyrroles I and III tilted toward the proximal side and
binds is due to potential steric clashes between the R- pyrroles II and IV tilted toward the distal side. Distortion
isomer and surrounding protein groups. As shown in Fig. from planarity permits an expansion of the inner core of
4, the ethyl substituent at the quaternary carbon is oriented pyrrole nitrogens to accommodate the predominantly high-
towards the distal pocket. The R-isomer would have the spin iron atom in CcP at room temperature [37]. Some
51
ethyl group oriented toward Trp in the distal pocket distortion of heme planarity is observed in most heme
˚resulting in contacts less than 2 A without a substantial proteins. Oxidized tuna cytochrome c has a saddle shaped
repositioning of the entire porphyrin. The oxo-substituent heme like CcP [38], while the heme in aquo-metmyoglobin
in the S-isomer is located toward the center of the protein is bowl shaped [39]. Model heme X-ray structures of
and away from the putative guaiacol binding site on the hexa-coordinate high-spin complexes show that the iron is
175
surface (see Fig. 5). Likewise, the modified heme edge is in the plane. The side chain of His and a water molecule
˚nearly 10 A from the molecular surface where little occupies the fifth and sixth coordinating positions, respec-
variance with WTCcP is found. Although the propionate tively (Fig. 4).

8. 642 C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643
reconstituted oxochlorin protein, and confirm that the
altered activity is not due to large structural changes in the
reconstituted hybrid.
5. Abbreviations
CcP cytochrome c peroxidase
COSY correlation spectroscopy
ddab dimethyldidodecylammonium bromide
EPR electron paramagnetic resonance
ESIMS electrospray ionization mass spec-
trometry.
Mp or mesopone 8,12,-diethyl-3,8,13,17-tetramethyl-7-
oxo-prophyrin-2,18-diproprionic acid
MpCcP mesopone-reconstituted CcP
MPD 2-methyl-2,4-pentanediol
NOE nuclear overhauser effect
WTCcP wild-type CcP
Fig. 5. Active site structure of MpCcP, showing the positioning of the
oxo-substituent towards the center of the protein, opposite the suggested
site of guaiacol binding.
Acknowledgements
This research was supported by the National Science
Another interesting feature of this high resolution struc- Foundation (P.J.F., CHE-0100774), the Petroleum Re-48
ture is that the distal Arg , a residue involved in the search Fund (P.J.F., PRF-31804-G3), Undergraduate Re-
formation and stabilization of compound I, which is in search Opportunities Program funding from the University
multiple conformations in WTCcP, is locked in one of California, Irvine and the National Institutes of Health
conformation as shown in the omit F 2 F difference mapo c (T.L.P., GM 42614). C.E.I. acknowledges a graduate
(Fig. 4). This conformation, which we have termed ‘in’, is fellowship from the UC Toxic Substances Research and
similar to what is found in compound I [40] and in the Teaching Program. M.C. acknowledges support from the
fluoride complex [41]. The ‘in’ conformation enables Arnold and Mabel Beckman Undergraduate Scholarship48
Arg to directly interact with ligands coordinated to the Fund.
heme iron or, in the case of the high-spin ferric complex,
to an ordered water molecule situated ligated to the iron at
˚a distance of 1.96 A. In this regard, the structural features
References
of MpCcP resemble that of the enzyme intermediate of
WTCcP with hydrogen peroxide rather than the resting
¨ ¨[1] V. Fulop, J.W.B. Moir, S.J. Ferguson, J. Hajdu, Cell 81 (1995)
state of the enzyme. 369–377.
[2] C.K. Chang, R. Timkovich, W. Wu, Biochemistry 25 (1986) 8447–
8453.
[3] S. Junemann, Biochim. Biophys. Acta 1321 (1997) 107–127.4. Conclusion
[4] J.J. Hill, J.O. Alben, R.B. Gennis, Proc. Natl. Acad. Sci. USA 90
(1993) 5863–5867.
Heme ring oxygenation has a variable effect on CcP [5] C.K. Chang, K.M. Barkigia, L.K. Hanson, J. Fajer, J. Am. Chem.
activity that is not readily explained by the altered chemi- Soc. 108 (1986) 1352–1354.
cal properties of the cofactor or structural changes in the [6] C.K. Chang, L.K. Hanson, P.F. Richardson, R. Young, J. Fajer, Proc.
Natl. Acad. Sci. USA 78 (1981) 2652–2656.protein. The reactivity of MpCcP with native substrate
[7] K.M. Barkigia, C.K. Chang, J. Fajer, M.W. Renner, J. Am. Chem.
cytochrome c is virtually unchanged from the wild type,
Soc. 114 (1992) 1701–1707.191
and the ready formation of the putative Trp radical [8] Y. Liu, M.D. Ryan, Inorg. Chim. Acta 225 (1994) 57–66.
intermediate is confirmed by EPR spectroscopy of the [9] Y. Liu, C. DeSilva, M.D. Ryan, Inorg. Chim. Acta 258 (1997)
compound I state of the hybrid protein. The expected 247–255.
[10] S. Ozawa, E. Sakamoto, T. Ichikawa, Y. Watanabe, I. Morishima,increase in Fe-based reduction potentials would suggest an
Inorg. Chem. 34 (1995) 6362–6370.
increased activity with small molecule substrates, and is
[11] R. Dmello, S. Hill, R.K. Poole, Microbiol. UK 142 (1996) 755–763.
seen with guaiacol, but activity with ferrocyanide is only [12] V. Borisov, R. Gennis, A.A. Konstantinov, Biochem. Mol. Biol. Int.
slightly decreased. The results of a single crystal structural 37 (1995) 975–982.
determination offer the first structural parameters for a [13] J.H. Dawson, Science 240 (1988) 433–439.

9. C.E. Immoos et al. / Journal of Inorganic Biochemistry 91 (2002) 635–643 643
[14] J.E. Erman, L.B. Vitello, J.M. Mauro, J. Kraut, Biochemistry 28 [28] C.K. Chang, W. Wu, J. Org. Chem. 51 (1986) 2134–2137.
(1989) 7992–7997. [29] R.K. Pandey, M. Isaac, I. MacDonald, C.J. Medforth, M.O. Senge,
[15] S. Ozawa, Y. Watanabe, S. Nakashima, T. Kitagawa, I. Morishima, J. T.J. Dougherty, K.M. Smith, J. Org. Chem. 62 (1997) 1463–1472.
Am. Chem. Soc. 116 (1994) 634–641. [30] T.J. Dougherty, K.M. Smith, J. Org. Chem. 67 (1997) 1463–1472.
[16] L.A. Fishel, J.E. Villafranca, J.M. Mauro, J. Kraut, Biochemistry 27 [31] A.F. Nassar, J.M. Bobbit, J.D. Stuart, J.F. Rusling, J. Am. Chem.
(1987) 351–360. Soc. 117 (1995) 10986–10993.
[17] K. Choudhury, M. Sundaramoorthy, A. Hickman, T. Yonetani, E. [32] Z. Lu, Q. Huang, J.F. Rusling, J. Electroanal. Chem. 423 (1997)
Woehl, M.F. Dunn, T.L. Poulos, J. Biol. Chem. 269 (1994) 20239– 59–66.
20249. [33] M. Bayachou, R. Lin, W. Cho, P.J. Farmer, J. Am. Chem. Soc. 120
[18] S.L. Edwards, T.L. Poulos, J. Biol. Chem. 265 (1990) 2588–2595. (1998) 9888–9898.
[19] M. Sundaramoorthy, K. Choudhury, S.L. Edwards, T.L. Poulos, J. [34] S.A. Martinis, C. Sotiriou, C.K. Chang, S.G. Sligar, Biochemistry
Am. Chem. Soc. 113 (1991) 7755–7757. 28 (1989) 879–884.
[20] Z. Otwinoski, W. Minor, Methods Enzymol. 276 (1996) 307–326. [35] G.D. DePillis, B.P. Sishta, A.G. Mauk, P.R. Ortiz de Montellano, J.
[21] A.T. Brunger, P.D. Adams, G.M. Clore, W.L. DeLano, P. Gros, R.W. Biol. Chem. 266 (1991) 19334–19341.
Grosse-Kunstleve, J.S. Jiang, J. Kuszewski, M. Nilges, N.S. Pannu, [36] S.K. Wilcox, G.M. Jensen, M.M. Fitzgerald, D.E. Mcree, D.B.
R.J. Read, L.M. Rice, T. Simonson, G.L. Warren, Acta Crystallogr. Goodin, Biochemistry 35 (1996) 4858–4866.
D54 (1998) 905–921. [37] T. Iizuka, M. Kotani, T. Yonetani, J. Biol. Chem. 246 (1971)
[22] A.T. Brunger, Nature 355 (1992) 472–475. 4731–4736.
[23] D.J. Schuller, Acta Crystallogr. D52 (1996) 425–434. [38] T. Takano, R.E. Dickerson, J. Mol. Biol. 153 (1981) 79–94.
[24] T. Yonetani, H. Schleyer, A. Eherenberg, J. Biol. Chem. 241 (1966) [39] T. Takano, J. Mol. Biol. 110 (1977) 537–568.
3240. [40] V. Fulop, R.P. Phizackerley, S.M. Soltis, I.J. Clifton, S. Wakatsuki, J.
[25] R.M. Fowler, H.A. Bright, J. Res. Natl. Bur. Stand. 15 (1935) 493. Erman, J. Hajdu, S.L. Edwards, Structure 2 (1994) 201–208.
[26] V.P. Miller, G.D. DePillis, J.C. Ferrer, A.G. Mauk, P.R. Ortiz de [41] S.L. Edwards, T.L. Poulos, J. Kraut, J. Biol. Chem. 259 (1984)
Montellano, J. Biol. Chem. 267 (1994) 8936–8942. 12984–12988.
[27] J.M. Mauro, L.A. Fishel, J.T. Hazzard, T.E. Meyer, G. Tollin, M.A.
Cusanovich, J. Kraut, Biochemistry 27 (1994) 6243–6256.