Nitrite increases the enantioselectivity of sulfoxidation catalyzed by myoglobin derivatives in the presence of hydrogen peroxide

Article abstract of DOI:10.1016/j.tet.2004.06.097

The effect of nitrite in the sulfoxidation of organic sulfides catalyzed by myoglobin (Mb) in the presence of hydrogen peroxide has been investigated. A general improvement in enantioselectivity was found for the reaction catalyzed by horse heart metMb and a series of sperm whale metMb derivatives including the wild type protein, the active site mutants T67K Mb, T67R Mb, T67R/S92D Mb, and the T67K Mb derivative reconstituted with the modified prosthetic group protohemin-l- histidine methyl ester. Graphical Abstract

Full text of DOI:10.1016/j.tet.2004.06.097

Tetrahedron 60 (2004) 8153–8160
Nitrite increases the enantioselectivity of sulfoxidation catalyzed
by myoglobin derivatives in the presence of hydrogen peroxide
Vincenza Pironti,a Stefania Nicolis,b Enrico Monzani,b Stefano Colonnaa and Luigi Casellab,
*
a
` `
Istituto di Chimica Organica ‘Alessandro Marchesini’, Facolta di Farmacia, Universita di Milano, via Venezian 21, 20133 Milano, Italy
b
`
Dipartimento di Chimica Generale, Universita di Pavia, Via Taramelli 12, 27100 Pavia, Italy
Received 20 April 2004; revised 25 May 2004; accepted 17 June 2004
Available online 21 July 2004
Abstract—The effect of nitrite in the sulfoxidation of organic sulfides catalyzed by myoglobin (Mb) in the presence of hydrogen peroxide
has been investigated. A general improvement in enantioselectivity was found for the reaction catalyzed by horse heart metMb and a series of
sperm whale metMb derivatives including the wild type protein, the active site mutants T67K Mb, T67R Mb, T67R/S92D Mb, and the T67K
Mb derivative reconstituted with the modified prosthetic group protohemin-L- histidine methyl ester.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
site of Mb, the group of Watanabe obtained mutants that
exhibit significant catalytic turnover with high stereo-
selectivity in the sulfoxidation and epoxidation of sub-
stituted aromatic substrates.4a,6,10 A different strategy of Mb
modification involves replacement of the natural prosthetic
group with a synthetic hemin.4b,11
Recent studies show that many proteins can provide low
levels of activity in reactions different from those involved
in their normal biological function.1 Activity can be
improved by engineering the proteins, particularly in the
active site features. Such a strategy has become popular in
the case of metalloproteins, where the metal-binding sites
have been redesigned in different ways, to introduce
structural features which are specific to other proteins or
enzymes.2 Among these proteins is myoglobin (Mb), the
protein of storage and intracellular transfer of molecular
oxygen,3 which is widely used as a catalyst in peroxide-
dependent one-electron oxidation, such as oxidation4 and
nitration5 of phenolic substrates, and in reactions involving
formal two-electron oxidation, such as sulfoxidation of
organic sulfides and epoxidation of alkenes.6 The latter type
of reactions, termed peroxygenations, can proceed with a
certain degree of stereoselectivity. Both enantioselective
sulfoxidation7 and epoxidation8 are catalyzed much more
efficiently by chloroperoxidase, but in general, commercial
applications of peroxidases are limited by the high cost and
the low stability of the enzyme, due to the rapid inactivation
by the oxidant.9 For this reason, other proteins such as Mb
and its derivatives are explored as potential biocatalysts to
perform asymmetric transformations. Single or double
mutations can provide substantial contributions to the
optimization of the new activity: by engineering the active
We have recently found that during the catalytic cycle
of peroxidase-type reactions by Mb, nitrite, in the
presence of hydrogen peroxide, produces two powerful
nitrating and oxidizing species, nitrogen dioxide (NO$2)
and peroxynitrite (ONOOK), depending on nitrite
concentration.5 Herein we report on the effect of nitrite
in the oxidation of aromatic sulfides catalyzed by horse
heart Mb (hh Mb), sperm whale Mb (WT Mb), the three
active site mutants of the latter protein T67R Mb,12
T67K Mb13 and T67R/S92D Mb,14 and the mutant
T67K Mb reconstituted with the modified hemin
obtained by covalently linking an L-histidine methyl
ester residue to one of the propionate side chains of
protohemin IX, T67K-His Mb.4b,13 All the proteins are
utilized in their met (Fe3C) form. The aim of this
investigation is to assess whether the highly reactive
species generated in the presence of nitrite could
participate in the sulfoxidation catalytic cycle of Mb
and increase the efficiency and/or stereoselectivity of
the reaction. Indeed, a general improvement in enantio-
selectivity was obtained both for horse heart Mb and
sperm whale Mb mutants. In addition, we performed
NMR studies on Mb-substrate complexes with the aim
of gaining a picture of the interaction between the
sulfides and the protein.
Keywords: Asymmetric oxidation; Myoglobin; Nitrite; NMR;
Sulfoxidation.
* Corresponding author. Tel.: C39-382-507331; fax: C39-382-528544;
e-mail: bioinorg@unipv.it
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.06.097

8154
V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
2. Results and discussion
2.1. Effect of nitrite in the sulfoxidation reaction
The sulfoxidation of thioanisole (1) catalyzed by hh Mb in
the presence of hydrogen peroxide was chosen as a standard
reaction to optimize reaction conditions. The effects of
reaction time, temperature, amount of hydrogen peroxide,
and concentration of Mb on the enantioselectivity and
conversion of sulfoxide were taken into consideration in
these preliminary experiments. It was also established that
the non-catalyzed reaction, that is, the reaction carried out in
the absence of Mb by simply mixing the sulfide with
hydrogen peroxide (and nitrite when required) was
negligible in the optimized conditions (see Section 4). The
effect of nitrite concentration on sulfoxide yield and
enantiomeric excess was then studied in order to investigate
the eventual involvement of the two different active species
derived from nitrite in the mechanism of sulfoxidation.
Indeed, an improvement in enantioselectivity by nitrite was
observed and the concentration of nitrite that maximizes the
enantioselectivity was found to be 50 mM (Table 1). At this
relatively low [NOK2 ] the only active species derived from
nitrite which may be involved in the sulfoxidation is
nitrogen dioxide.5 On the other hand, the yield of sulfoxide
decreased in the presence of nitrite, indicating that part of
the hydrogen peroxide was consumed in the parallel,
unproductive oxidation of nitrite to nitrate.
The increase of enantioselectivity in the sulfoxidation
carried out in the presence of nitrite was confirmed with
the other substrates (Table 2). We chose sulfides carrying
a para substituent, methyl p-tolylsulfide (2), the iso-propyl
p-tolylsulfide (3), and methyl 2-pyridylsulfide (4). Different
reactions conditions were needed for compound 4 (see
Section 4), because this substrate is more difficult to oxidize
than the others, due to the electron withdrawing effect of the
heteroatom. For this reason, it was necessary to increase the
reaction time and add the oxidizing agent and nitrite in small
aliquots in order to protect the Mbs from degradation.
In order to exclude the possibility that nitrite acted in the
reaction through a simple salt effect, that is, by an
electrostatic interaction with the protein, the sulfoxidation
of 1 catalyzed by hh Mb was carried out in the presence of
sodium nitrate. These experiments were made at different
concentrations of nitrate and in every case the enantiomeric
excess did not change with respect to the reactions carried
out in the absence of nitrate.
2.2. Asymmetric sulfoxidation catalyzed by Mb mutants
In our previous studies, we prepared Mb mutants containing
Table 1. Effect of addition of nitrite in the sulfoxidation of thioanisole
catalyzed by hh Mb in the presence of hydrogen peroxide (the R enantiomer
is always the major sulfoxide product)
Conversion (%)
ee (%)
[NOK2 ] (M)
39G4
15G2
24G1
20G1
28G1
10G2
21G2
29G3
22G1
5G1
—
0.01
0.05
0.10
0.20

V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
8155
Table 4. Sulfoxidation of methyl 2-pyridyl sulfide catalyzed by Mb mutants
and hydrogen peroxide in the presence of 50 mM nitrite (the R enantiomer
is always the major sulfoxide product)
a basic residue in the heme distal pocket, by replacement of
Thr67 with either Arg (T67R Mb)12 or the more flexible Lys
residue (T67K Mb),13 and a double mutant where, in
addition to the Thr67Arg mutation, the proximal Ser92 was
substituted with an Asp residue (T67R/S92D Mb).14 These
modifications aimed at introducing into the Mb active site
those amino acid residues which are critical for the
activation of hydrogen peroxide in peroxidases.4b In
addition, we prepared a derivative of T67K Mb recon-
stituted with protohemin-L-histidine methyl ester (T67K-
His Mb).13 The latter hemin contains only one free
carboxylate group and therefore, reconstitution of Mb
with the modified cofactor involves the loss of the
interaction with one of the propionate groups which
stabilize heme binding to the protein. This causes a
relaxation of the protein fragment around the heme which
was found to increase the accessibility of donor molecules
and phenolic substrates to the active site.4b
Mb
Conversion (%)
ee (%)
Horse heart
WT
T67R
12G1
39G4
76G5
40G4
24G2
19G3
33G3
0
T67K-His
heme peroxidases resulted in a ‘hydrogen abstraction
oxygen-rebound’ mechanism; the actual oxygen transfer
reagent could be one of the high-valent oxoferryl inter-
mediates known as compound I (P$C FeIV]O, where P$C
indicates a porphyrin cation radical) and compound II
(FeIV]O), depending on the nature of the enzyme used.15
For Mbs, unlike peroxidases, the high-valent oxoferryl
species resulting from oxidation of Fe3C by hydrogen
peroxide cannot be differentiated, because instead of a
porphyrin cation radical the initial species contains the
radical localized on protein residues.16 However, the
possibility to stabilize a ‘compound I-like’ species for the
Mb mutants obtained upon replacement of the distal His64
residue, allowed Watanabe and co-workers to investigate
the direct sulfide-induced reduction of different compound I
species and compare the mechanism of sulfoxidation of
these proteins with that of horseradish peroxidase (HRP).17
Two distinct pathways were found: the reduction of HRP
compound I by thioanisoles proceeds via electron transfer in
the protein cage (followed by oxygen rebound from
compound II), whereas the reduction of H64S Mb
compound I proceeds via direct oxygen transfer (the
intermediate compound II being not involved). Compound
I was observed as the catalytic species for the peroxygenase
activity of all His64 Mb mutants, whereas only compound II
was observed with WT Mb.18
All Mb mutants shared with hh Mb the positive effect of
nitrite in the sulfoxidation of thioanisole (Table 3). Among
the Mb mutants studied here, the best enantioselective
catalyst was found to be T67R Mb, since it afforded the
highest increase of enantiomeric excess with respect to WT
Mb in the presence of nitrite (from 15 to 51%). The yield of
sulfoxide was also significantly improved. The highest
conversion to sulfoxide was actually obtained with the
mutant T67K Mb, but in this case the enantiomeric excess
did not increase with respect to WT Mb. The low
asymmetric induction is possibly due to the fact that the
lysine residue at position 67 is not rigid enough to control
the preferential formation of one sulfoxide enantiomer.
Apparently, the guanidinium group in the arginine side
chain exerts stronger steric control on the binding of the
sulfide to the protein, that leads to a better stereodifferentia-
tion in the oxygenation process.
Considering that all the proteins investigated here contain
the His64 residue, both the oxoferryl intermediates
produced in the catalytic cycle, that we indicate as.
MbFeIV]O and MbFeIV]O, are involved in the sulfoxida-
tion. The catalytic scheme can be summarized as follows:
The T67R Mb derivative gives higher enantioselectivity and
yield also in the oxidation of substrate 4 (Table 4). In this
case, the reactions were only carried out in the presence of
nitrite, since the enhancement of enantioselectivity by
nitrite in the oxidation of 4 was already observed for the
reaction catalyzed by hh Mb (Table 2). With both the
substrates, the T67K-His Mb derivative produced racemic
sulfoxides: the increased protein mobility around the heme
is clearly not advantageous for the immobilization of the
substrate and results in the formation of achiral sulfoxides.
$
MbFe3C CH2O2 / MbFeIV]O CH2O
(1)
(2)
(3)
(4)
$MbFeIV]O CS/MbFe3C CSO
$MbFeIV]O CS/½MbFeIV]O CS$C
½MbFeIV]O CS$Cꢀ/MbFe3C CSO
ꢀ
2.3. Mechanism of sulfoxidation
Mechanistic investigations of oxygen transfer reactions by
Table 3. Sulfoxidation of thioanisole catalyzed by Mb mutants and hydrogen peroxide in the presence of 50 mM nitrite (the R enantiomer is always the major
sulfoxide product)
[NOK2 ]Z0 M
[NOK2 ]Z0.05 M
Mbs
Conversion (%)
ee (%)
Conversion (%)
ee (%)
Horse heart
WT
T67R
T67K
T67R/S92D
T67K-His
39G4
30G4
61G4
27G3
1G0.5
12G1
10G2
13G2
20G2
24G1
22G2
54G4
75G5
25G2
40G3
29G1
15G1
51G2
15G3
9
10G2
a
0
0
a
The low conversion does not allow a reliable estimate of the enantiomeric excess.

8156
V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
½MbFeIV]O CS$Cꢀ/MbFeIV]O CS$C
(5)
(6)
(7)
Nitrite could participate in the sulfoxidation reaction also
through the oxidant species NO$2 produced by reaction (8).
Nevertheless, by reacting thioanisole with nitrogen dioxide
gas, we did not observe appreciable formation of the
sulfoxide during the reaction time of the catalytic
experiments. As a result, nitrite acts only as a quencher of
the intermediate MbFeIV]O in the present system. It is
worth nothing that increasing the nitrite concentration above
50 mM, where the Mbs can generate peroxynitrite,5 the
enantiomeric excess in the sulfoxidation reaction drops
(Table 1). Probably, the formation of ONOOK implies a
decreased reactivity of $MbFeIV]O with the sulfide and, at
the same time, the highly reactive ONOOK readily diffuses
into the bulk solution, where its reaction with the sulfide
produces racemic sulfoxide.
MbFeIV]O CS/MbFe3C CS$C ðslowÞ
2S$C CH2O/S CSO C2HC
where MbFe3C is the initial metMb form, S indicates the
sulfide, SO the sulfoxide and S$C the sulfur cation radical,
respectively.
$
The first protein intermediate, MbFeIV]O, can convert
sulfide to sulfoxide with a certain degree of stereoselectivity
by a direct O-transfer reaction (formally a two-electron
$
process), reaction (2). MbFeIV]O may also form a sulfur
cation radical by one-electron oxidation in the protein cage
(indicated with square brackets) according to reaction (3).
The latter reaction converts the protein into the second
intermediate, MbFeIV]O, containing a S$C cation radical
in the active site. This species can evolve according to two
competitive pathways: it can give the oxygen rebound to
afford the sulfoxide with a certain degree of stereoselectivity
according to reaction (4), or the sulfenium radical can
diffuse from the protein cage according to reaction (5). The
following reduction of MbFeIV]O by another molecule of
sulfide produces with low efficiency (see below) a sulfur
cation radical, reaction (6). Finally, the dismutation of sulfur
cation radicals in the bulk of the solution by reaction (7)
gives the racemic product. Therefore, the stereoselectivity
in the sulfoxidation depends on the fraction of Mb that
reacts via $MbFeIV]O (reaction (2)) or via oxygen rebound
from MbFeIV]O in the protein cage (reaction (4)).
The formation of racemic sulfoxide in the reaction catalyzed
by the T67K-His Mb both in the absence and in the presence
of nitrite (see Table 3) confirms the effect of nitrite
according to the proposed mechanism. In fact, if the oxygen
$
transfer from MbFeIV]O (or from MbFeIV]O in the
protein cage) is not stereoselective, then quenching of
MbFeIV]O by nitrite according to reaction (8) cannot
increase the enantiomeric excess in the sulfoxidation.
Considering the fast reaction of nitrite with the intermediate
MbFeIV]O, also the rate of the sulfoxidation, and conse-
quently the conversion into product, should increase in the
presence of nitrite. However, a fraction of hydrogen
peroxide is involved in the unproductive oxidation of nitrite
to nitrate according to the reaction:
NOK2 CH2O2 /NOK3 CH2O
(9)
Reaction (6) is a slow, low efficiency, process. In fact,
treating the MbFeIV]O intermediate of hh Mb with
different amount of thioanisole did not significantly affect
the rate of disappearance of the MbFeIV]O bands; that is,
the formation of metMb and Mb dimers through self
oxidation of the active species is faster than reaction with
the sulfide.
which is also catalyzed by Mb. As a result, the yields of
sulfoxide in the presence of nitrite are smaller than those
obtained without nitrite.
1
2.4. NMR relaxation measurements and H NMR
spectroscopy
In contrast, nitrite efficiently reduces the MbFeIV]O inter-
mediate to the native state with a second order rate constant
kZ(18.6G0.6) MK1 sK1 according to the reaction:5
To gain an understanding of the protein–substrate inter-
action, as this is the basis for the stereoselective effects in
the sulfoxidation reaction, we performed H NMR relaxa-
1
MbFeIV]O CNOK2 /MbFe3C CNO$2
(8)
tion time measurements on substrate 4 in the presence of a
representative set of proteins, including the best enantio-
selective catalyst (T67R Mb), the catalyst that produces
racemic sulfoxide (T67K-His Mb), and WT Mb, for com-
parison purposes. In these experiments, the paramagnetic
contribution to relaxation by the high spin Fe3C center of
the protein can be exploited to get an estimate of the
distances of the protons of the bound substrate from the iron
atom. The study had to be restricted to substrate 4, because it
is the only sulfide sufficiently soluble in aqueous buffer;
attempts to obtain NMR relaxation data with other
substrates gave unreliable data due to the very limited
range of concentrations attainable. As shown by the data
collected in Table 5, the iron-proton distances for protein
This process is faster than reaction (6). Furthermore, the
reduction rate of MbFeIV]O by nitrite is not affected by the
presence of thioanisole in the reaction mixture (see Section
4), indicating that the sulfide does not inhibit reaction (8) to
a significant extent.
According to this mechanism, nitrite efficiently competes
with sulfide for reduction of MbFeIV]O. As a result, in the
presence of this anion the importance of reaction (6) is
reduced and, along with this, the contribution of the
pathway that produces the sulfoxide through the stereo-
chemically unproductive coupling of sulfur cation radicals.
The protein returning to its native form can be involved in
another catalytic cycle. Therefore, the effect of nitrite is to
increase the fraction of protein that proceeds via two-
electron oxidation of the substrate and thereby increase also
the stereoselectivity of sulfoxidation.
˚
bound 4 were in the range of 6.9–7.3 A for WT Mb, 5.8–
˚
˚
6.2 A for T67R Mb, and 6.5–7.0 A for T67K-His Mb. The
similarity in the iron-proton distances for the substrate
protons in each protein complex clearly indicates that the
sulfide maintains a certain degree of mobility even when

V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
8157
Table 5. Substrate proton relaxation times and iron–proton distances for protein–substrate 4 complexes of Mb derivatives, in deuterated 0.2 M phosphate buffer
pD 7.5, 25 8C
Proton (ppm)
WT Mb
T67R Mb
T67K-His Mb
˚
r (A)
˚
r (A)
˚
r (A)
T1M (s)
T1M (s)
(9.3G0.9)!10K4
(9.8G0.4)!10K4
T1M (s)
H-3 7.29
H-4 7.66
H-5 7.10
H-6 8.27
CH3 2.47
(2.6G0.1)!10K3
(3.2G0.2)!10K3
(3.5G0.2)!10K3
(3.1G0.1)!10K3
(2.7G0.1)!10K3
6.9
7.2
7.3
7.1
7.0
5.8
5.9
6.1
6.0
6.2
(1.9G0.1)!10K3
(1.8G0.2)!10K3
(2.1G0.1)!10K3
(2.3G0.1)!10K3
(2.8G0.2)!10K3
6.6
6.5
6.7
6.8
7.0
(1.16G0.06)!10K3
(1.09G0.05)!10K3
(1.33G0.06)!10K3
bound to the protein active site. The distance values agree
with a disposition of methyl 2-pyridyl sulfide partially
inside the distal cavity of the Mbs, as previously found for
the binding to Mb of p-cresol, which has comparable size.12
The sulfide can approach the heme in T67R Mb more
closely than in WT Mb, suggesting that this can be a key
feature for the enhancement of the enantioselectivity of the
sulfoxidation reaction (see Table 4). In the complex
between substrate 4 and T67K-His Mb, the iron-proton
distances are larger than those observed for T67R Mb.
Moreover, the data account for a different orientation of the
substrate inside the cavity of T67R Mb and T67K-His with
respect to WT Mb, the aromatic protons being more inside
the heme cavity and the methyl group more outside in the
former cases. We can speculate that the basic residue (Arg
or Lys) present in the two mutants contributes to the binding
by hydrogen bonding to the sulfur atom or the pyridyl
nitrogen atom of the substrate. The interaction with the
sulfur atom, however, would be the only polar interaction
established by T67R Mb with thioanisole, which gives the
best enantioselectivity in the sulfoxidation. The larger
mobility of the polypeptide chain of T67K-His Mb around
the heme would prevent the relative immobilization of the
sulfides, which results in racemic products (Tables 3 and 4).
resonances of the propionate-7 protons Ha (from 72.7 to
0
71.0 ppm) and Ha (from 31.0 to 31.9 ppm), and for methyl-1
(from 51.3 to 52.0 ppm), but here also methyl-5 is affected
(from 83.8 to 84.5 ppm). The selective perturbation of the
NMR signals accounts for a disposition of both the sulfides
at least partially into the heme distal cavity, spanning the
space above the substituents at positions 1, 7 and 8 of the
porphyrin, and is in agreement with the range of distances of
the substrate protons from the iron center obtained in the
relaxation rate measurements.
3. Conclusion
We have shown that nitrite can be used as a reagent to
improve the enantioselectivity of the sulfoxidation of
organic sulfides catalyzed by several Mb derivatives in
the presence of hydrogen peroxide. A model for the
interaction between the substrates and the proteins has
been obtained through NMR experiments. Although the
sulfoxidation of thioanisole catalyzed by the double mutants
H64D/V68A and H64D/V68S of Mb6b,10c has been reported
to occur with enantioselectivities higher than those found
here, the effect of nitrite is significant and may find other
applications in catalytic oxidation reactions.
In order to complete the analysis of the interaction between
the Mb derivatives and the sulfides, the paramagnetic H
1
NMR spectra of the proteins were studied in the presence of
excess sulfide. The downfield region of the NMR spectrum
of horse heart Mb is compared in Figure 1 with the spectrum
obtained after the addition of substrate 1 to the protein. The
assignment of the resonances of hh Mb are made on the
basis of the data previously reported.19 The most notable
differences can be found in the chemical shifts of the signals
0
of the heme propionate-7 protons, Ha and Ha : while the
former undergoes an upfield shift of 1.7 ppm, the latter
shifts 0.5 ppm downfield. Actually, this propionate group is
localized in the solvent accessible side of the heme moiety.
A slight shift of 0.3 ppm upfield can be detected also for the
methyl group in position 1 of the porphyrin, probing the
ability of thioanisole to deeply enter into the heme distal
cavity.
The addition of substrate 4 to a solution of hh Mb produces
appreciable changes in several of the downfield paramag-
netic signals (Fig. 2); this is probably due to polar effects
established by methyl 2-pyridyl sulfide within the protein
active site. The largest shifts are again observed for the
Figure 1. Downfield region of 400 MHz 1H NMR spectrum of hh Mb
(w0.1 mM) in deuterated 0.2 M phosphate buffer pD 7.5 in the presence of
substrate 1 (0.5 mM) at 25 8C (lower trace), compared with the protein
spectrum in the absence of substrate (upper trace). The assignment of the
peaks, according to Ref. 19, is shown.

8158
V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
containing hh Mb or the Mb mutants (5.0 mM). H2O2
(0.50 mM) and NaNO2 (0.05 M), when required, were
added in 100 min in 10 aliquots at 10 min intervals. After
5 min additional time the reaction mixture was extracted
with dichloromethane (3!10 ml) and the organic phase
dried with Na2SO4. The solvent was removed under reduced
pressure to give the crude product, which was then analyzed
by HPLC.
The HPLC analyses were performed with a Merck–Hitachi
L-7100 pump and a DAD 1050 HP detector on Daicel chiral
columns OD and OB, at a flow rate of 0.8 ml/min. Elution
conditions were as follows: column OD, 20% isopropanol,
80% hexane for 1; column OB, 20% isopropanol, 80%
hexane for 2; column OB, 15% isopropanol, 85% hexane for
3 and 4. Readings were made at 230 nm for 1–3 and at
210 nm for 4. On Chiralcel OB the (S)-sulfoxides eluted
first, whereas on Chiralcel OD the elution order was the
opposite.21 Standard curves prepared using synthetic
sulfoxides were used for quantitative analysis and the
value of % conversion and enantiomeric excess were
determined on the basis of the peak areas of HPLC traces.
Figure 2. Downfield region of 400 MHz 1H NMR spectrum of hh Mb
(w0.1 mM) in deuterated 0.2 M phosphate buffer pD 7.5 in the presence of
substrate 4 (12 mM) at 25 8C (lower trace), compared with the protein
spectrum in the absence of substrate (upper trace). The assignment of the
peaks, according to Ref. 19, is shown.
4. Experimental
4.1. Materials
4.3. Reactions of MbFeIV]O with thioanisole and with
nitrite
The hh MbFeIV]O intermediate was prepared by incubat-
ing metMb (4 mM) in 0.2 M phosphate buffer, pH 7.5 with
2 equiv. H2O2 for about 15 min, until the Soret band shifted
from 410 to 420 nm and stabilized at this wavelength. The
reaction with 1 was monitored observing the return of the
Soret band from 420 to 410 nm with time, after the addition
of the substrate at different concentrations (up to 0.5 mM).
The spontaneous evolution of MbFeIV]O to metMb was a
slow process and was not influenced by thioanisole, since an
appreciable increment in the rate of this process in the
presence of the sulfide was not observed.
The substrates 1 and 2 were from Aldrich, while 3 and 4
were synthesized as previously reported.20 The sulfoxides
were synthesized from the corresponding sulfide by
oxidation with sodium metaperiodate. Horse heart Mb was
obtained from Sigma as a lyophilized sample. The protein
derivatives WT Mb, T67R Mb, T67K Mb and T67R/S92D
Mb were produced, expressed and purified as previously
reported.12–14 The synthesis of hemin-L-histidine methyl
ester and the reconstitution of T67K Mb with the modified
hemin were performed as previously reported.13
The concentration of Mb solutions was determined from the
extinction coefficients of the met forms in 100 mM
phosphate buffer, pH 6.0, as follows: hh Mb, 1.88!
105 MK1 cmK1 at 408 nm;3 WT Mb, 1.57!105 MK1 cmK1
at 410 nm;12 T67K Mb, 1.52!105 MK1 cmK1 at 408 nm;13
T67R Mb, 1.49!105 MK1 cmK1 at 410 nm;12 T67R/S92D
Mb, 1.61!105 MK1 cmK1 at 408 nm;14 T67K-His Mb,
1.33!105 MK1 cmK1 at 410 nm.13 The other reagents were
obtained from commercial sources at the best grade
available.
Moreover, the observed first-order rate constants for
reduction of MbFeIV]O by nitrite (followed spectro-
photometrically under pseudo-first order conditions as
previously reported)5 were not affected by the presence of
thioanisole (0.5 mM) in the reaction mixture.
4.4. Reaction of NO$2 with thioanisole
NO$2 was obtained by air oxidation of NO$. 200 ml of 1 atm
gaseous NO$2 (a slight excess with respect to the amount that
gives a final concentration of 1 mM) were bubbled through a
gas-tight syringe into a solution of substrate 1 (0.50 mM) in
4 ml of 0.2 M phosphate buffer, pH 7.5. After 5 min, the
solution was extracted with CH2Cl2, the organic phase was
dried under vacuum and then analyzed by HPLC as reported
above (see Section 4.2).
4.2. Enzymatic sulfoxidation
Method A—The sulfides (1–3) (0.50 mM) were reacted, at
25 8C, under stirring, in 4 ml of 10 mM phosphate buffer,
pH 7.5, containing hh Mb or the Mb mutants (5.0 mM) and
H2O2 (0.50 mM). When required, NaNO2 (0.010–0.20 M)
in the usual phosphate buffer was added. After 5 min the
reaction mixtures were extracted with dichloromethane (3!
10 ml) and the combined organic phase dried with Na2SO4.
The solvent was removed under reduced pressure to give the
crude product, which was then analyzed by HPLC.
1
4.5. NMR relaxation measurements and H NMR
spectroscopy
The T1 relaxation time for the protons of substrate 4 in the
presence of variable amounts of WT Mb, T67R Mb, or
T67K-His Mb were determined at 25 8C with a Bruker
AVANCE 400 NMR spectrometer operating at
400.13 MHz, using the standard inversion recovery
Method B—Sulfide 4 (0.50 mM) was reacted, at 25 8C,
under stirring in 4 ml of 10 mM phosphate buffer, pH 7.5,

V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
8159
method.22 The solutions of 4 were prepared in deuterated
0.2 M sodium phosphate buffer pD 7.5 and contained
different amount of the proteins. In order to eliminate
interferences by metal impurities, a small amount of EDTA
was added to the solutions. The concentrations employed
were the following: [4]Z12.0 mM and [WT Mb]Z0–6 mM;
[4]Z14.0 mM and [T67R Mb]Z0–2 mM; [4]Z14.5 mM
and [T67K-His Mb]Z0–4 mM. The relaxation rate of the
protons of the substrate molecules interacting with the Mbs,
Fukuzumi, S.; Watanabe, Y. Biochemistry 2003, 42, 10174–
10181.
7. (a) Colonna, S.; Gaggero, N.; Manfredi, A.; Casella, L.;
Gullotti, M.; Carrea, G.; Pasta, P. Biochemistry 1990, 29,
10465–10468. (b) Colonna, S.; Gaggero, N.; Casella, L.;
Carrea, G.; Pasta, P. Tetrahedron: Asymmetry 1992, 3, 95–106.
(c) Casella, L.; Gullotti, M.; Ghezzi, R.; Poli, S.; Beringhelli,
T.; Colonna, S.; Carrea, G. Biochemistry 1992, 31, 9451–9459.
(d) van Deurzen, M. P. J.; Groen, B. W.; van Rantwijk, F.;
Sheldon, R. A. Biocatalysis 1994, 10, 247–255. (e) van
Deurzen, M. P. J.; Remkes, I. J.; van Rantwijk, F.; Sheldon,
R. A. J. Mol. Catal. A: Chem. 1997, 117, 329–337. (f)
Dembitsky, V. M. Tetrahedron 2003, 59, 4701–4720.
8. (a) Colonna, S.; Gaggero, N.; Casella, L.; Carrea, G.; Pasta, P.
Tetrahedron: Asymmetry 1993, 4, 1325–1330. (b) Dexter,
A. F.; Lakner, F. J.; Campbell, R. A.; Hager, L. P. J. Am.
Chem. Soc. 1995, 117, 6412–6413. (c) Lakner, F. J.; Hager,
L. P. J. Org. Chem. 1996, 61, 3923–3925. (d) Lakner, F. J.;
Hager, L. P. Tetrahedron: Asymmetry 1997, 8, 3547–3550.
9. (a) Shevelkova, A. N.; Ryabov, A. D. Biochem. Mol. Biol.
1996, 39, 665–670. (b) Baynton, K. J.; Bewtra, K. J.; Biswas,
N.; Taylor, K. E. Biochim. Biophys. Acta 1994, 1206, 272–
278. (c) van Rantwijk, F.; Sheldon, R. A. Trends Biotechnol.
2001, 19, 73–80.
T1b, was calculated from experimental relaxation rate, T1obs
through the equation:23
,
ꢀ
ꢁ
1
1
1
E0
1
Z
K
C
(10)
T1obs
T1b T1f KD CS0 T1f
where T1f is the T1 value for free substrate, E0 and S0 are the
initial protein and substrate concentrations, respectively,
and KD is the dissociation constant for the Mb-substrate
complex. The major contribution to the T1b value is the
paramagnetic contribution (T1M), which is correlated to the
distance (r) of the nucleus from the Fe3C center according
to the Solomon–Bloembergen equation (assuming an
electron relaxation time ts of 5!10K11 s).12,24,25 The KD
values are not known and were neglected in the present
calculations; however, the magnitude of the error associated
with this approximation is identical for all the substrate
protons. It is possible to estimate for r a maximum error of
10% assuming a dissociation constant equal to the substrate
concentration.
10. (a) Ozaki, S.; Matsui, T.; Watanabe, Y. J. Am. Chem. Soc.
1996, 118, 9784–9785. (b) Ozaki, S.; Matsui, T.; Watanabe, Y.
J. Am. Chem. Soc. 1997, 119, 6666–6667. (c) Kato, S.; Yang,
H.; Ueno, T.; Ozaki, S.; Philips, G. N.; Fukuzumi, S. M.;
Watanabe, Y. J. Am. Chem. Soc. 2002, 124, 8506–8507.
11. Hayashi, T.; Hisaeda, Y. Acc. Chem. Res. 2002, 35, 35–43.
12. Redaelli, C.; Monzani, E.; Santagostini, L.; Casella, L.;
Sanangelantoni, A. M.; Pierattelli, R.; Banci, L. ChemBio-
Chem 2002, 3, 226–233.
The paramagnetic proton NMR spectra of hh Mb were
recorded at 25 8C on solutions of the protein (w0.1 mM) in
deuterated 0.2 M sodium phosphate buffer, pD 7.5. The
interaction of hh Mb with the sulfides was studied by
recording NMR spectra of the Mb solution containing
substrate 1 (w0.5 mM) or 4 (12 mM). Spectra were
recorded acquiring 5K scans, with a 80000 Hz spectral
window, and suppressing the water signal by presaturation
for 0.3 s.
`
13. Roncone, R.; Monzani, E.; Labo, S.; Sanangelantoni, A.M.;
Casella, L. Submitted for publication.
14. Roncone, R.; Monzani, E.; Murtas, M.; Battaini, G.; Pennati,
A.; Sanangelantoni, A. M.; Zuccotti, S.; Bolognesi, M.;
Casella, L. Biochem. J. 2004, 717–724.
15. (a) Perez, U.; Dunford, H. B. Biochim. Biophys. Acta 1990,
1038, 98–104. (b) Perez, U.; Dunford, H. B. Biochemistry
1990, 29, 2757–2763. (c) Baciocchi, E.; Lanzalunga, O.;
Malandrucco, S. J. Am. Chem. Soc. 1996, 118, 8973–8974. (d)
van Rantwijk, F.; Sheldon, R. A. Curr. Opin. Biotechnol. 2000,
11, 554–564.
Acknowledgements
This work was supported by a PRIN Project of the Italian
MIUR. We thank Giuseppe Celentano for HPLC analyses.
16. (a) Fenwick, C. W.; English, A. M. J. Am. Chem. Soc. 1996,
118, 12236–12237. (b) Degray, J. A.; Gunther, M. R.;
Tschirret-Guth, R. A.; Ortiz de Montellano, P. R. J. Biol.
Chem. 1997, 272, 2359–2362. (c) Sugiyama, K.; Highet,
R. J.; Woods, A.; Cotter, R. J.; Osawa, Y. Proc. Natl. Acad.
Sci. USA 1997, 94, 796–801. (d) Gunther, M. R.; Tschirret-
Guth, R. A.; Witowska, H. E.; Fann, Y. C.; Barr, D. P.; Ortiz de
Montellano, P. R.; Mason, R. P. Biochem. J. 1998, 330, 1293–
1299. (e) Tsukuhara, K.; Kiguchi, K.; Matsui, M.; Kubota, N.;
Arakawa, R.; Sakurai, T. J. Biol. Inorg. Chem. 2000, 5, 765–
773.
References and notes
1. O’Brien, P. J.; Herschlag, D. Chem. Biol. 1999, 6, R91–0.
2. Lu, Y.; Berry, S. M.; Pfister, T. D. Chem. Rev. 2001, 101,
3047–3080.
3. Antonini, E.; Brunori, M. Hemoglobin and Myoglobin in their
Reactions with Ligands; North-Holland: Amsterdam, 1971.
4. (a) Watanabe, Y.; Ueno, T. Bull. Chem. Soc. Jpn 2003, 76,
1309. (b) Roncone, R.; Monzani, E.; Nicolis, S.; Casella, L.
Eur. J. Inorg. Chem. 2004. 2203–2213.
17. Matsui, T.; Ozaki, S.; Watanabe, Y.; Fukuzumi, S. J. Am.
Chem. Soc. 1999, 121, 9497–9502.
18. Ozaki, S.; Matsui, T.; Roach, M. P.; Watanabe, Y. Coord.
Chem. Rev. 2000, 198, 39–59.
5. Nicolis, S.; Monzani, E.; Roncone, R.; Gianelli, L.; Casella, L.
Chem. Eur. J. 2004, 10, 2281–2290.
19. Lloyd, E.; Burk, D. L.; Ferrer, J. C.; Maurus, R.; Doran, J.;
Carey, P. R.; Brayer, G. D.; Mauk, A. G. Biochemistry 1996,
35, 11901–11912.
6. (a) Ozaki, S.; Yang, H.; Matsui, T.; Goto, Y.; Watanabe, Y.
Tetrahedron: Asymmetry 1999, 10, 183–192. (b) Yang, H.-J.;
Matsui, T.; Ozaki, S.; Kato, S.; Ueno, T.; Phillips, G. N.;
20. (a) Kelly, D. R.; Knowles, C. J.; Mahdi, J. G.; Taylor, I. N.;

8160
V. Pironti et al. / Tetrahedron 60 (2004) 8153–8160
Wright, M. A. Tetrahedron: Asymmetry 1996, 7, 365–368. (b)
Dawood, K. M.; Higashiya, S.; Hou, Y.; Fuchigami, T. J. Org.
Chem. 1999, 64, 7935–7939.
23. Banci, L.; Bertini, I.; Luchinat, C.; Nuclear and electron
relaxation The Magnetic Nucleus-Unpaired Electron
Coupling in Solution; VCH: Weinheim, 1991.
21. Colonna, S.; Gaggero, N.; Casella, L.; Carrea, G.; Pasta, P.
Tetrahedron: Asymmetry 1992, 3, 95–106.
24. Solomon, I. Physiol. Rev. 1955, 99, 559–565.
25. Bloembergen, N. J. Chem. Phys. 1957, 27, 572–573.
22. Vold, R. L.; Waugh, J. S.; Klein, M. P.; Phelps, D. E. J. Chem.
Phys. 1968, 48, 3831–3832.