New activities of a catalytic antibody with a peroxidase activity
Formation of Fe(II)–RNO complexes and stereoselective oxidation of sulfides
Re
´
my Ricoux, Edyta Lukowska, Fabio Pezzotti and Jean-Pierre Mahy
Laboratoire de Chimie Bioorganique et Bioinorganique, Institut de Chimie Mole
´
culaire et des Mate
´
riaux d’Orsay,
Universite
´
Paris-Sud XI, Orsay, France
In order to estimate the size of the cavity remaining around
the heme of the 3A3–microperoxidase 8 (MP8) hemo-
abzyme, the formation of 3A3–MP8–Fe(II)-nitrosoalkane
complexes upon oxidation of N-monosubstituted hydroxyl-
amines was examined. This constituted a new reaction for
hemoabzymes and is the first example of fully characterized
Fe(II)–metabolite complexes of antibody–porphyrin. Also,
via a comparison of the reactions with N-substituted hyd-
roxylamines of various size and hydrophobicity, antibody
3A3 was confirmed to bring about a partial steric hindrance
on the distal face of MP8. Subsequently, the influence of
the antibody on the stereoselectivity of the S-oxidation of
sulfides was examined. Our results showed that MP8 alone
and the antibody–MP8 complex catalyze the oxidation of
thioanisole by H
2
O
2

where the imidazole side-chain of histidine 18 acts as a
proximal ligand of the iron atom. A set of six monoclonal
antibodies was thus obtained: the best peroxidase activity –
that found with the complex of MP8 and one of those
antibodies, 3A3 – was characterized by a k
cat
/K
m
value of
2 · 10
6
M
)1
Æmin
)1
, the best ever reported for an antibody–
porphyrin complex [2]. Active-site topology studies sugges-
ted that the binding of MP8 occurred through interactions
of its carboxylate substituents with amino acids of the
antibody, and that the protein provided a partial steric
hindrance of the distal face of the heme [2]. In addition,
it was shown recently that 3A3–MP8 was a more efficient
catalyst for the nitration of phenol by NO
2
–
/H
2
O
2
than

´
Paris-Sud XI, 91405, Orsay cedex, France.
Fax: + 33 1 69 15 72 81, Tel.: + 33 1 69 15 74 21,
E-mail:
Abbreviations: CcP, cytochrome c peroxidase; CH
3
COOEt, ethyl-
acetate; CiP, Coprinus cinereus peroxydase; CPO, chloroperoxidase;
HRP, horseradish peroxidase; KLH, keyhole limpet hemocyanin;
LPO, lactoperoxidase; mCPBA, meta-chloroperbenzoic acid; MP8,
microperoxidase 8; MPO, myeloperoxidase; NOS, nitric acid
synthase; RNO, Fe(II)–nitrosoalkane complex; tBuOH, tert-butyl
alcohol; tBuOOH, tert-butyl hydroperoxide.
Enzymes: catalase (EC 1.11.1.6); horseradish peroxidase, myelo-
peroxidase, lactoperoxidase (EC 1.11.1.7); chloroperoxidase
(EC 1.11.1.10); cytochrome c peroxidase (EC 1.11.1.6).
(Received 8 December 2003, revised 21 January 2004,
accepted 6 February 2004)
Eur. J. Biochem. 271, 1277–1283 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04032.x
[8], as catalysts, and the S configuration in the presence
of horseradish peroxidase (HRP) [9], Coprinus cinereus
peroxydase (CiP) [7], cytochrome c peroxidase (CcP) [10]
and soybean peroxygenase [11]. The S-oxidation of organic
sulfides by peroxidases could involve two types of mech-
anisms (Scheme 1). The first mechanism is a Ôone-step
oxygen transfer mechanismÕ, with an oxygen atom being
directely transferred from Compound I to the sulfur atom
of the organic sulfide; the second mechanism is the Ôtwo-
step oxygen-transfer mechanismÕ which involves a radical–
cation intermediate.

O
2
occurs with a 45%
enantiomeric excess in favour of the R isomer. This
constitutes the highest enantiomeric excess reported to date
for the oxidation of sulfides catalyzed by porphyrin–
antibody complexes.
Materials and methods
Preparation of MP8
MP8 was prepared by sequential peptic and tryptic digestion
of horse-heart cytochrome c(Sigma), as described previously
[12]. The heme content was determined using the pyridine
chromogen method [12]. The purity of the sample was greater
than 97%, based on MALDI-TOF mass spectrometry.
Preparation of monoclonal antibodies
MP8 was covalently attached to keyhole limpet hemo-
cyanin (KLH) and to BSA, using glutaraldehyde as a
coupling agent, in 1
M
bicarbonate buffer, pH 9.5, according
to Tresca et al. [13]. The conjugates were then purified by
column chromatography on Biogel P10. Hapten–protein
ratios were determined spectrophotometrically using a molar
absorption coefficient value (e)of1.49· 10
5
M
)1
Æcm
)1
at

under acidic conditions generated the corresponding 1-p-
chlorophenyl-2-nitropropene [18], which was then reduced
into N-substituted hydroxylamine using a lithium-alumin-
ium hydride, as described previously [19].
Reaction of MP8 and 3A3–MP8 with N-monosubstituted
hydroxylamines
Forty microlitres of a 10
)2
M
solution of N-monosubsti-
tuted hydroxylamine in CH
3
OH was added to a cuvette
containing 1 mL of 0.86 l
M
MP8–Fe(III) or 3A3–MP8
complex [obtained by preincubation of 0.86 l
M
MP8–
Fe(III) with 2 l
M
antibody 3A3 for 1 h at room tempera-
ture in 0.1
M
NaCl/P
i
(PBS), pH 7.4]. The evolution of
the UV-visible spectrum of the solution was monitored, as
Scheme 1. Mechanisms of oxygen transfer reactions catalyzed by per-
oxidases.

formed after 10 min were calculated using a m
M
absorption
coefficient (e ¼ 7.87 m
M
)1
Æcm
)1
at 254 nm) in the differ-
ence spectrum between the sulfide and the corresponding
sulfoxide.
Stereoselective oxidation of thioanisole. Standard incu-
bations (total volume, 0.5 mL) were performed at room
temperature in Tris buffer (0.1
M
, pH 7.5) containing
thioanisole (100 l
M
) and the catalyst, either 0.3 l
M
MP8
alone or 3A3–MP8 prepared by preincubation of 0.3 l
M
MP8 with 0.87 l
M
antibody for 1 h at room temperature.
An oxidant – H
2
O
2

i
,
pH 7.4, with antibody 3A3 (2 l
M
),ledtonewcomplexes,3a
and 3b, respectively, characterized by an absorption spec-
trum similar to those observed for the MP8–Fe(II)–RNO
complexes, with aborption maxima at approximately 413
and 530 nm (Table 1) [20]. The reactivity of these new
complexes was very similar to that of the MP8–Fe(II)–RNO
complexes and that of other already reported hemoprotein–
Fe(II)–RNO complexes [20], in that (a) they were stable
for 2 h in 0.1
M
NaCl/P
i
, pH 7.4, in the presence of 1 m
M
sodium dithionite, and (b) conversely, they were rapidly
destroyed upon the addition of 100 l
M
potassium ferri-
cyanide [Fe(CN)
6
K
3
], with regeneration of the 3A3–MP8–
Fe(III) complex. This strongly suggested a 3A3–MP8–
Fe(II)–RNO structure for these new complexes 3a and 3b.
Such a structure was confirmed by the following result, that

Table 1. UV-visible characteristics of microperoxidase 8 (MP8) and 3A3–MP8–RNO complexes and kinetic constants for their formation by reaction
of MP8 and 3A3–MP8 with N-substituted hydroxylamines.
Fe(II)–RNO complex. R ¼
UV-visible
MP8
b
3A3–MP8
c
MP8
b
3A3-MP8
b
k
max
(nm), e (m
M
)1
Æcm
)1
)
a
k(min
)1
)
d
C
50
(l
M
)

–kt
), using Kaleidagraph.
e
The RNHOH concentration which
leads to 50% conversion of MP8 or 3A3–MP8 into the corresponding Fe(II)–RNO complex.
Ó FEBS 2004 N- and S-oxidations catalyzed by a hemoabzyme (Eur. J. Biochem. 271) 1279
above results validate the use of hemoabzymes as a
convenient model for hemoproteins used in toxicology
and pharmacology, such as cytochrome P450, peroxidases
and nitric oxide synthase (NOS). It is probable that the
mechanism of formation of these complexes is similar to
that described for the formation of the MP8–Fe(II)–RNO
complexes and for the Fe(II)–RNO complexes of hemo-
proteins (Scheme 2). It should involve, first, a one-electron
reduction of the Fe(III) into Fe(II) by the monosubstituted
hydroxylamine to give the RNHOH
•+
radical cation. A
second, one-electron oxidation could then be achieved using
O
2
which, after losing two protons, should produce the
nitrosoalkane RNO that binds to MP8–Fe(II).
The values of the molar extinction coefficients at
413 nm for the 3A3–MP8–Fe(II)–RNO complexes have
been calculated according to Ricoux et al.[20](Table1).
From Table 1, it is clear that (a) the e-values depend on
the nature of the R substituent of the hydroxylamine and
(b) for both hydroxylamines, the e-valuesarelowerforthe
3A3–MP8–Fe(II)–RNO complexes than for their MP8–

)1
Æcm
)1
). Overall,
the minor changes observed when comparing the spectral
characteristics of the 3A3–MP8–Fe(II)–RNO complexes
with those of the MP8–Fe(II)–RNO complexes (i.e.
almost no shift and a slightly lower absorbance of the
soret band) have already been observed when inserting
MP8 into 3A3 [2]. They are consistent with the insertion
of the MP8–Fe(II)–RNO complex into a hydrophobic
pocket with no change of the Fe(II) spin state and no
replacement of any of the two axial ligands of the iron,
His18 or RNO, by an amino acid side-chain of the
antibody protein.
Binding site topology of antibody 3A3
Figure 1 shows the changes in the concentration of the
Fe(II)–RNO complex, formed upon addition of RNHOH
to MP8–Fe(III) or 3A3–MP8–Fe(III), as a function of time.
From this figure, it appears that the formation of the Fe(II)–
RNO complexes follows pseudo first-order kinetics, and
that the formation rate of MP8–Fe(II) or 3A3–MP8–
Fe(II)–RNO complexes depends on the hydroxylamine
structure (Fig. 1, Table 1). Interestingly, in both instances,
Fe(II)–RNO complexes derived from the smaller aliphatic
hydroxylamine (1), formed more rapidly than those derived
from the more bulky aromatic hydroxylamine (2). Indeed,
rate constants of 0.77 ± 0.04 min
)1
and 0.72 ± 0.02

M
. However, the concentra-
tion necessary to convert 50% of MP8 or 3A3–MP8
(0.86 l
M
) into the Fe(II)–RNO complex (C
50
)alsodepen-
ded on the hydroxylamine structure (Table 1). Indeed,
whereas very similar concentrations of hydroxylamine 1
and 2 were needed to convert 50% of 0.86 l
M
MP8 into
the corresponding Fe(II)–RNO complex (300 ± 5 l
M
and
285 ± 5 l
M
, respectively) (Table 1), a much higher con-
centration of hydroxylamine 2 (C
50
¼ 565 ± 5 l
M
)than
of hydroxylamine 1 (C
50
¼ 360 ± 5 l
M
) was needed to
convert 50% of 3A3–MP8 (0.86 l

complex formed at a given time and C
max
is the maximum concen-
tration of Fe(II)–RNO complex formed. MP8, microperoxidase 8;
RNO, Fe(II)–nitrosoalkane complex.
Scheme 2. Mechanism of the formation of 3A3–MP8Fe(II)–RNO
complexes and oxidation of these complexes by potassium ferricyanide.
1280 R. Ricoux et al. (Eur. J. Biochem. 271) Ó FEBS 2004
alkyl group, N-isopropyl-hydroxylamine-1, is more reactive
than hydroxylamine 2,whichissubstitutedbyanelectro-
attractive p-chlorophenyl group. Indeed, it leads to the
highest rate constant and the lowest concentration necessary
to convert them into an Fe(II)–RNO complex (C
50
), with
either MP8 or 3A3-MP8 (Fig. 1, Table 1). Second, with
both hydroxylamines, a decrease in the reaction rate, as well
as an increase in the C
50
value, are observed with the 3A3–
MP8 complex, when compared with MP8 alone. These
phenomena are particularly important in the case of the
more bulky and hydrophobic N-substituted hydroxylamine
2, as the reaction rate decreases by a factor of 1.7 while the
C
50
value increases by a factor of 2 (Table 1). This suggests
that, although the antibody 3A3 does not prevent the
binding of a ligand, such as nitrosoalcane, to the iron of
MP8, it brings a partial steric hindrance on the distal face of

)was
added to a solution of 84 l
M
thioanisole and 0.2 l
M
MP8 in
0.1
M
Tris buffer, pH 7.5, at room temperature. The
concentration of product formed after 10 min was calcula-
ted as described in the Materials and methods. The reaction
was quenched by the addition of excess Na
2
SO
3
, and the
organic products were then extracted with CH
3
COOEt and
analyzed by GC. The only product formed, with a 2.5%
yield, was the corresponding sulfoxide that was identified by
comparison with an authentic sample (Fig. 2, Table 2).
The involvement of the iron atom of MP8 in the catalysis
was indicated by the 100% inhibition of the reactions
performed in the presence of 100 m
M
CN
–
(data not
shown). Indeed, CN

as H
2
O
2
, mCPBA or t-BuOOH, in the presence of 10%
various organic solvents (methanol, CH
3
CN, t-BuOH). The
reactions were initiated by adding the oxidant, and the
concentrations of product formed after 10 min were calcu-
lated as described in the Materials and methods. The values
thus obtained are compared in Table 2. It first appeared
that H
2
O
2
was the best oxidant for the sulfoxidation of
thioanisole, as it produced the best yield in sulfoxide,
regardless of the solvent used. When tBuOOH was used in
the buffer alone, no oxidation was observed. However, in
the presence of organic solvents, sulfoxide was produced,
but in a lower yield than when using H
2
O
2
as an oxidant.
Finally, whatever the conditions, no sulfoxide was formed
when mCPBA was used as an oxidant, which confirmed
that the reaction occurred through a peroxidase Ôtwo-step
oxygen-transfer mechanismÕ, involving a radical–cation

microperoxidase 8 (MP8)
as the catalyst.
Oxidant
PhSOCH
3
(%)
Buffer
alone
+ 10%
methanol
+ 10%
CH
3
CN
+ 10%
t-BuOH
H
2
O
2
2.5 4.7 7.2 10.0
t-BuOOH – 2.1 6.2 2.8
mCBPA – – – –
Ó FEBS 2004 N- and S-oxidations catalyzed by a hemoabzyme (Eur. J. Biochem. 271) 1281
Stereoselective S-oxidation of thioanisole
As the above results showed that the best system for the
S-oxidation of thioanisole associated H
2
O
2

that could
lead to racemic sulfoxide. The reaction was then quenched
by the addition of excess Na
2
SO
3
, and the organic products
were then extracted with CH
3
COOEt and analyzed by GC
and HPLC, as described in the Materials and methods. The
results shown in Table 3 show that the antibody–MP8
complex is a more efficient catalyst than MP8 alone, either
with or without 5% tBuOH, and generates sulfoxide yields
of 30% and 49%, respectively, under these conditions,
whereas MP8 alone generates yields of 10% and 23%,
respectively, under the same conditions. The yields did not
exceed 49%, even in the best case, because an oxidative
degradation of the catalyst occurred. This was shown by a
progressive disappearance, in its absorption spectrum, of
the soret band at 396 nm that is characteristic of the heme
moiety. This degradation was less important in the case of
the antibody–MP8 catalyst, which showed that the antibody
protected the heme against oxidative degradation and led to
higher yields in sulfoxide. In addition, whereas almost no
enantiomeric excess is observed in the presence of MP8
alone, an important enantiomeric excess is observed with
3A3–MP8 used as a catalyst, with the best value of 45%
obtained in favor of the R enantiomer in the presence of 5%
tBuOH. These results confirm the important role of the

porphyrin–antibody (SN 37.4), which produced the
S-enantiomer sulfoxide with a 43% enantiomeric excess
[25]. Our results thus validate the use of the hemoabzyme
3A3–MP8 as a catalyst for the selective oxidation of
interesting substrates such as alkanes and alkenes.
References
1. Mahy, J.P., Desfosses, B., de Lauzon, S., Quilez, R., Desfosses, B.,
Lion, L. & Mansuy, D. (1998) Hemoabzymes. Different strategies
for obtaining artificial hemoproteins based on antibodies. Appl.
Biochem. Biotechnol. 75, 103–127.
2. Ricoux, R., Sauriat-Dorizon, H., Girgenti, E., Blanchard, D. &
Mahy, J.P. (2002) Hemoabzymes: towards new biocatalysts for
selective oxidations. J. Immunol. Methods 269, 39–57.
3. Ricoux, R., Girgenti, E., Sauriat-Dorizon, H., Blanchard, D. &
Mahy, J.P. (2002) Regioselective nitration of phenol induced by
catalytic antibodies. J. Protein Chem. 21, 473–477.
4. Damani, L.A. (1989) Sulphur-Containing Drugs and Related
Organic Compounds, Vol. 1, Ellis Horwood, Chichester.
5. Hofstra, A.H. & Uetrecht, J.P. (1993) Myeloperoxidase-mediated
activation of xenobiotics by human leukocytes. Toxicology 82,
221–242.
6. Colonna, S., Gaggero, N., Manfredi, A., Casella, L., Gullotti, M.,
Carrea, G. & Pasta, P. (1990) Enantioselective oxidations of
sulphides catalyzed by chloroperoxidase. Biochemistry 29, 10465–
10468.
7. Tuynman, A., Vink, M.K., Dekker, H.L., Schoemaker, H.E. &
Wever, R. (1998) The sulphoxidation of thioanisole catalysed by
Table 3. Yield and enantiomeric excess for the S-oxidation of thioani-
sole by H
2

1282 R. Ricoux et al. (Eur. J. Biochem. 271) Ó FEBS 2004
lactoperoxidase and Coprinus cinereus peroxidase: evidence for an
oxygen-rebound mechanism. Eur. J. Biochem. 258, 906–913.
8. Capeillere-Blandin, C., Martin, C., Gaggero, N., Pasta, P., Carrea,
G. & Colonna, S. (1998) Sulphoxidation reaction catalysed by
myeloperoxidase from human leucocytes. Biochem. J. 335, 27–33.
9. Colonna, S., Gaggero, N., Carrea, G. & Pasta, P. (1992) Horse-
radish peroxidase catalysed sulfoxidation is enantioselective.
J.Chem.Soc.Chem.Com.4, 357–358.
10. Miller, V.P., DePillis, G.D., Ferrer, J.C., Mauk, A.G. & Ortid de
Montellano, P.R. (1992) Monooxygenase activity of cytochrome c
peroxidase. J. Biol. Chem. 267, 8936–8942.
11. Ble
´
e, E. & Schuber, F. (1989) Mechanism of S-oxidation reaction
catalysed by a soybean hydroperoxide dependent oxygenase.
Biochemistry 28, 4962–4967.
12. Aron, J., Baldwin, D.A., Marques, H.M., Pratt, J.M. & Adams,
P.A. (1986) Hemes and hemoproteins 1: preparation and analysis
of heme containing octapeptide (Microperoxidase-8) and identi-
fication of the monomeric form in aqueous solution. J. Inorg.
Biochem. 27, 227–243.
13. Tresca, J.P., Ricoux, R., Pontet, M. & Engler, R. (1995) Com-
parative activity of peroxidase–antibody conjugates with periodate
and glutaraldehyde coupling according to an enzyme immuno-
assay. Ann. Biol. Clin. 53, 227–231.
14. Ko
¨
hler, G. & Milstein, C. (1975) Continuous culture of fused
cells secreting antibodies of predefined specificity. Nature 256,

metric sulfoxidation in organic solvents versus in water.
Biotechnol. Bioeng. 70, 353–357.
24. De Gioia, L., Ghibaudi, E., Laurenti, E., Salmona, M. & Ferrari,
R.P. (1996) A theoretical three-dimensional model for lactoper-
oxidase and eosinophil peroxidase, built on the scaffold of the
myeloperoxidase. J. Biol. Inorg. Chem. 1, 476–485.
25. Shai, S. & Keinan, E. (1999) Antibody-metalloporphyrin catalytic
assembly mimics natural oxidation enzymes. J. Am. Chem. Soc.
121, 8978–8982.
Ó FEBS 2004 N- and S-oxidations catalyzed by a hemoabzyme (Eur. J. Biochem. 271) 1283