Tài liệu Báo cáo khoa học: Models and mechanisms of O-O bond activation by cytochrome P450 A

REVIEW ARTICLE

Models and mechanisms of O-O bond activation by cytochrome P450

A critical assessment of the potential role of multiple active intermediates

in oxidative catalysis

Peter Hlavica

Walther-Straub-Institut fu¨r Pharmakologie und Toxikologie der LMU, Mu¨nchen, Germany

Cytochrome P450 enzymes promote a number of oxidative

biotransformations including the hydroxylation of unacti￾vated hydrocarbons. Whereas the long-standing consensus

view of the P450 mechanism implicates a high-valent iron￾oxene species as the predominant oxidant in the radicalar

hydrogen abstraction/oxygen rebound pathway, more

recent studies on isotope partitioning, product rearrange￾ments with
radical clocks, and the impact of threonine

mutagenesis in P450s on hydroxylation rates support the

notion of the nucleophilic and/or electrophilic (hydro)

peroxo-iron intermediate(s) to be operative in P450 catalysis

in addition to the electrophilic oxenoid-iron entity; this may

contribute to the remarkable versatility of P450s in substrate

modification. Precedent to this mechanistic concept is given

by studies with natural and synthetic P450 biomimics. While

the concept of an alternative electrophilic oxidant necessi￾tates C-H hydroxylation to be brought about by a cationic

insertion process, recent calculations employing density

functional theory favour a
two-state reactivity scenario,

implicating the usual ferryl-dependent oxygen rebound

pathway to proceed via two spin states (doublet and quar￾tet); state crossing is thought to be associated with either an

insertion or a radicalar mechanism. Hence, challenge to

future strategies should be to fold the disparate and some￾times contradictory data into a harmonized overall picture.

Keywords: (hydro)peroxo-iron; iron-oxene; O2-activation;

P450 biomimics; P450.

Introduction

Cytochrome P450 (P450 or CYP) enzymes (EC 1.14.14.1),

a superfamily of b-type hemoproteins found in organisms

from all domains of life [1], are major catalysts in the

oxidative biotransformation of a structural diversity of

endogenous and exogenous compounds [2]. While the

general chemistry of substrate hydroxylation has been

assessed on a broad basis, the specific problem of dioxygen

activation during P450 cycling is still the most important

and intriguing one in the area of P450 research. Here, the

need for an active oxidant capable of insertion into

unactivated C-H bonds in hydrocarbons and related

compounds has extensively captured the imagination of

biochemists owing to the unfavourable thermodynamics of

the dissociation event [3]. Early views of such a mechanism

focused on an oxygen insertion pathway promoted by an

electrophilic, high-valent iron-oxo species (compound I) [4].

This hypothesis was soon supplanted by the
hydrogen

abstraction/oxygen rebound concept implicating the exist￾ence of radical intermediates, as developed on the basis of

the well-known chemical properties of peroxidases and

porphyrin model systems [5,6]. The mechanistic details of

oxygen transfer have been addressed elsewhere [7,8].

Mounting evidence provided during the past decade

suggests that hydroxylation reactions are more complex

than previously anticipated, and are not compatible with

the idea of a single reaction pathway. The picture began to

cloud when the application of ultrafast
radical clocks

to time the oxygen-rebound step disclosed the amounts of

rearranged products not to correlate with the radical

rearrangement rate constants [9]. Moreover, the use of a

probe that could distinguish between radical and cationic

species hinted at the interference of cationic rearrangements,

predicting the hydroxylation to occur via an insertion

reaction in place of abstraction and recombination [9]. The

former process thus necessitated the insertion into a C-H

bond of the elements of OH+, implying that the ultimate

electrophilic oxidant was either hydroperoxo-iron or iron￾complexed hydrogen peroxide [10]. In addition, examina￾tion of the oxidative deformylation of cyclic aldehydes as a

model for the demethylation reaction mediated by steroido￾genic P450s strongly favoured nucleophilic attack on the

Correspondence to P. Hlavica, Walther-Straub-Institut fu¨r Pharmak￾ologie und Toxikologie, Goethestr. 33, D-80336 Mu¨nchen, Germany.

Fax: +49 89 218075701, Tel.: +49 89 218075706,

E-mail: [email protected]

Abbreviations: TSR, two-state reactivity; KIE, kinetic isotope effects;

Hb, haemoglobin; Mb, myoglobin; HO, heme oxygenase; PDO,

phthalate dioxygenase; TDO, toluene dioxygenase; NDO, naphtha￾lene 1,2-dioxygenase; PMO, putidamonooxin; BLM, bleomycin;

NOS, nitric oxide synthases.

Enzymes: Cytochrome P450 (EC 1.14.14.1); NADPH-cytochrome

P450 oxidoreductase (EC 1.6.2.4); heme oxygenase (EC 1.14.99.3);

phthalate oxygenase reductase (EC 1.18.1); phthalate dioxygenase

(EC 1.14.12.7); toluene dioxygenase (EC 1.14.12.11); naphthalene

1,2-dioxygenase (EC 1.14.12.12); putidamonooxin (EC 1.14.99.15);

nitric oxide synthases (EC 1.14.13.39).

(Received 29 July 2004, revised 27 September 2004,

accepted 28 September 2004)

Eur. J. Biochem. 271, 4335–4360 (2004)
FEBS 2004 doi:10.1111/j.1432-1033.2004.04380.x

substrates by an iron-peroxo intermediate [11]. The sum of

these findings points at the involvement of more than one

active oxidant in the diverse types of P450-catalyzed

substrate processing [12–15].

The goal of the present perspective is to provide a critical

update of several aspects of the current state of biochemistry

relating to the apparently complex machinery of dioxygen

activation, which is considered to possibly implicate mul￾tiple oxygenating species in P450 catalysis. Emphasis will be

put on the evaluation of comparative studies with non-P450

hemoproteins, nonheme metalloenzymes as well as bio￾mimetic model systems to discuss the
multiple oxidant vs.

the
two-state reactivity theory.

Iron-oxene acting as an electrophilic oxidant

in P450-catalyzed hydroxylations

The consensus mechanism for hydrocarbon hydroxylation

by P450 enzymes involves hydrogen atom abstraction from

the hydrocarbon by a high-valent iron-oxo species, best

described as an O ¼ Fe(IV) porphyrin p-cation radical,

followed by homolytic substitution of the alkyl radical thus

formed in the so-called
oxygen rebound step [5–8]

(Scheme 1). Using CYP2B isoforms as the catalysts, radical

collaps was demonstrated to occur at highly variable rates

exceeding those of the gross molecular motions of many

enzyme-bound substrates and depending on the stereo￾chemical specificities of the compounds to be acted upon

[16,17]. Reduction of ferric P450 to the ferrous state sets the

stage for dioxygen binding, the event that commits the

hemoprotein to the step-by-step production of the active

oxidant (Scheme 2). Association of dioxygen with ferrous

microsomal CYP1A2 [18], certain CYP2B isoforms [19–21],

and CYP2C3 [18] to yield hexacoordinate low-spin com￾plexes has been shown to be characterized by absorption

bands around 420 and 557 nm in the absolute spectra and

broad maxima at about 440 and 590 nm in the difference

spectra. Similar optical perturbations were also observed

upon O2 binding to so-called class I P450s, comprising

mitochondrial and bacterial isozymes such as CYP11A1

[22–24] and CYP101 [25,26], respectively. The rapid initial

step in molecular oxygen activation by both class I and class

II P450s, as measured at varying temperatures, usually

exhibits monophasic kinetic behaviour, with the second￾order rate constants ranging from 0.58 to 8.41 · 106 M)1

Æs

)1

[18,20,24,25]. Interestingly, the presence of certain substrates

such as aromatic amines appears to favour homotropic

cooperativity in dioxygen binding to P450s: using liver

microsomal samples from untreated rabbits, the O2 satura￾tion kinetics for acetanilide 4-hydroxylation have been

reported to bear sigmoidal character corresponding to a Hill

interaction coefficient, n, of 2.2 [27]. Similar experiments

with N-alkyl arylamines gave concave upward double￾reciprocal plots of velocity vs. O2 concentration, from which

n could be calculated to have a value of 2.0–2.1 [28,29].

Apparent cooperativity in dioxygen association was found

to be highly sensitive to changes in hydrogen ion concen￾tration and was most pronounced at physiological pH,

whereas CO, acting as a positive effector, abolished

autoactivation at all pH values examined (Fig. 1) [30]. In

view of the well-known microheterogeneity of several rabbit

liver P450s [31], the amine-induced cooperativity in O2

complexation has been argued to involve the equilibrium

between multiple, kinetically distinct protein conformations

[32]. Alternatively, the oligomeric nature of P450 [33] might

offer the possibility of substrate–specific subunit inter￾actions, as has been proposed for the fractional saturation

of hemoglobin by dioxygen [34].

Results from resonance Raman spectroscopy [35] and

Mo¨ssbauer studies [36] with microbial CYP101 indicate that

Scheme 1. Rebound mechanism for P450-catalyzed hydroxylations.

Reproduced from [6] with permission.

Fe(III) Fe(II) Fe(III)

O

O

Fe(III)

O

O

Fe(III)

O

O

H

Fe(IV)

O

Fe(III) Fe(III)

HO

OH

Fe(IV)

HO OH O

+ H+

+ H+ + H+ H2O

O2

peroxo-iron

nucleophilic oxidant

hydroperoxy-iron

(inserts OH+

?)

oxo-ion, low spin

(inserts O)

spin inversion

iron-complexed hydrogen peroxide

(inserts OH+

?)

oxo-ion, high spin

(abstracts H+

)

e e

Scheme 2. The putative iron-oxygen inter￾mediates in P450 and their possible roles as

oxidants. Data collated from [10,15] with

permission.

4336 P. Hlavica (Eur. J. Biochem. 271)
FEBS 2004