Free radical trapping properties of several ethyl-substituted derivatives of 5-ethoxycarbonyl-5-methyl-1-pyrroline N-oxide (EMPO)

Article abstract of DOI:10.1016/j.bmc.2007.02.025

The spin trapping behavior of several ethyl-substituted EMPO derivatives, cis- and trans-5-ethoxycarbonyl-3-ethyl-5-methyl-pyrroline N-oxide (3,5-EEMPO), 5-ethoxycarbonyl-4-ethyl-5-methyl-pyrroline N-oxide (4,5-EEMPO), cis- and trans-5-ethoxycarbonyl-5-ethyl-3-methyl-pyrroline N-oxide (5,3-EEMPO), and 5-ethoxycarbonyl-5-ethyl-4-methyl-pyrroline N-oxide (5,4-EEMPO), toward a series of different oxygen- and carbon-centered radicals, is described. Considerably different stabilities of the superoxide adducts (ranging from about 12 to 55 min) as well as the formation of other radical adducts were observed.

Full text of DOI:10.1016/j.bmc.2007.02.025

Bioorganic & Medicinal Chemistry 15 (2007) 2827–2836
Free radical trapping properties of several ethyl-substituted
derivatives of 5-ethoxycarbonyl-5-methyl-1-pyrroline
N-oxide (EMPO)
Klaus Stolze,a,* Natascha Rohr-Udilova,b Thomas Rosenau,c
Andreas Hofingerc and Hans Nohla
aMolecular Pharmacology and Toxicology Unit, Department of Natural Science,
University of Veterinary Medicine Vienna, Veterina¨rplatz 1, A-1210 Vienna, Austria
bInstitute of Cancer Research, Department of Medicine I, Medical University of Vienna, Borschkegasse 8a, A-1090 Vienna, Austria
cDepartment of Chemistry, University of Natural Resources and Applied Life Sciences (BOKU),
Muthgasse 18, A-1190 Vienna, Austria
Received 6 December 2006; revised 6 February 2007; accepted 13 February 2007
Available online 15 February 2007
Abstract—The spin trapping behavior of several ethyl-substituted EMPO derivatives, cis- and trans-5-ethoxycarbonyl-3-ethyl-5-
methyl-pyrroline N-oxide (3,5-EEMPO), 5-ethoxycarbonyl-4-ethyl-5-methyl-pyrroline N-oxide (4,5-EEMPO), cis- and trans-5-eth-
oxycarbonyl-5-ethyl-3-methyl-pyrroline N-oxide (5,3-EEMPO), and 5-ethoxycarbonyl-5-ethyl-4-methyl-pyrroline N-oxide
(5,4-EEMPO), toward a series of different oxygen- and carbon-centered radicals, is described. Considerably different stabilities of
the superoxide adducts (ranging from about 12 to 55 min) as well as the formation of other radical adducts were observed.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
trans-isomer of 3,5-EDPO was found to be very stable
(t1/2 ca. 45 min), whereas the superoxide adduct from
the cis-isomer is rather short-lived (t1/2 ca. 11 min). The
respective stereoisomers from 4,5-EDPO and the other
investigated compounds could, however, not be sepa-
rated into the respective diastereomers using conven-
tional chromatographic procedures. In our attempts to
optimize the stability of the spin adducts and also to pro-
vide a whole range of spin traps of different lipophilic
properties we also investigated a series of pentyl- and
phenyl-substituted EMPO derivates with comparatively
poor spin trapping properties.3 In this paper, we describe
the spin trapping behavior of several ethyl-substituted
EMPO derivatives (cis-3,5-EEMPO, trans-3,5-EEMPO,
cis-/trans-4,5-EEMPO, cis-5,3-EEMPO, trans-5,3-EEM-
PO, and cis-/trans-5,4-EEMPO) toward a series of differ-
ent oxygen- and carbon-centered radicals.
Previous studies have shown that variation of the sub-
stituents in EMPO-derived spin traps has a great influ-
ence on the stability of their superoxide spin adducts,1–5
especially if a methyl group is present in position 3 or 4
of the pyrroline ring.1–3 In the case of 3,5-EDPO (5-eth-
oxycarbonyl-3,5-dimethyl-pyrroline N-oxide) two differ-
ent diastereomeric forms of the spin trap, which are
chemically distinct due to the presence of two symmetric
carbon centers in the pyrroline ring, could be separated
by conventional column chromatography. The respective
cis- and trans-forms exhibited considerably different spin
trapping properties, in contrast to observations made by
Tsai et al.5 using enantiomers of 5-t-butoxycarbonyl-5-
methyl-1-pyrroline N-oxide. After the addition of super-
oxide, an additional asymmetric center is created, result-
ing in the formation of two spectroscopically different
spin adducts from either of these compounds. The pre-
dominant superoxide spin adduct formed from the
2. Results
2.1. Structure of the spin traps
Keywords: EPR; Spin trapping; Superoxide; EMPO derivatives; Lipid-
derived radicals.
The spin traps synthesized and investigated within the
present study can be considered as 5-ethoxycarbonyl-
*
Corresponding author. Tel.: +43 1 25077 4406; fax: +43 1 25077
4490; e-mail: Klaus.Stolze@vu-wien.ac.at
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.02.025

2828
K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
5-methyl-1-pyrroline-N-oxide (EMPO) derivatives car-
rying an ethyl substituent at either C-3, C-4, or C-5.
We would like to use the abbreviation EEMPO for these
compounds, with two numbers indicating the position of
the ethyl substituent and the methyl group, respectively.
Thus, 3,5-EEMPO has an ethyl substituent at C-3 and a
methyl substituent at C-5 according to the numbering
used (see Fig. 1).
ethoxycarbonyl group was largely invariant against sub-
stitution influences: the carbonyl appeared at about
169–170 ppm throughout, and the ethoxy signals were
found at approximately 13.8 and 62.0 ppm.
In contrast to this structure, the resonances of the
methyl and ethyl substituents were much more depen-
dent on substitution position and configuration. The
4-methyl substituent was found at 14–15 ppm, the
3-methyl group at 18–19 ppm. The 5a-methyl group
was generally found around 21 ppm (with the exception
of the trans-4,5-EEMPO derivative), where a high-field
shift to 14.8 ppm appeared. The methyl resonance of
5-ethyl substituents, being in the range of 7–8 ppm,
experienced a down-field shift by 4 ppm for 3-ethyl
and 4-ethyl derivatives. The methylene resonances of
the ethyl groups were less sensitive to structural changes.
Four substitution patterns, 3-ethyl-5-methyl, 4-ethyl-5-
methyl, 5-ethyl-3-methyl, and 5-ethyl-4-methyl, were
realized. Due to the presence of two stereo centers at
C-5 and C-3 (or C-4, respectively) the synthesis afforded
a mixture of two diastereomers for each compound with
a certain substitution pattern, which in the case of 3,5-
EEMPO and 5,3-EEMPO could be separated and puri-
fied. Depending on whether the ethoxycarbonyl group
and the alkyl substituent at C-3/C-4 are located on the
same side of the heterocyclic plane or at different sides,
the compounds were designated cis or trans. The alcohol
component of the ester was kept constant as ethyl
throughout the current compound series.
The proton spectra exhibited a typical resonance pat-
tern, too. H-2 exhibited coupling constants of 2.2–
2.3 Hz and appeared expectedly as doublet or triplet
for 3-substituted and 4-substituted derivatives, respec-
tively. The appearance of H-2 as a triplet—not a doublet
of doublets—indicated similar coupling constants to
geminal H-3 protons. 3a-CH3 substituents were found
at 1.25 ppm for the cis-derivative and at 1.20 ppm for
the trans-compound, 4a-CH3 groups resonated more
up-field at about 1.18 ppm (trans-compound) and
1.09 ppm (cis-derivative). The 5a-CH3 group resonated
at 1.58 ppm in 3,5,5-trisubstituted pyrrolines with no
difference between cis- and trans-isomers, whereas in
the 4,5,5-trisubstituted counterparts there was an influ-
ence of the configuration, with the 5a-CH3-resonance
of the cis-isomer (1.70 ppm) being shifted about
0.1 ppm down-field relative to the trans-isomer.
The structural identity of the novel spin traps was
confirmed by ESI Q-TOF HR-MS (Table 1), FTIR
(Table 2), UV–vis, and NMR spectroscopy. A complete
1
set of H, HAH correlated, 13C, HMQC, and HMBC
spectra was recorded for each compound, which allowed
C
for a complete signal assignment in both the 1H and 13
domain. Carbon NMR confirmed the presence of a het-
erocyclic pyrroline ring. The resonances for C-2 were
found around 136–140 ppm according to the respective
substitution and configuration pattern, those for C-5
ranged between 79 and 85 ppm. C-5 exhibited a typical
down-field shift by about 3 ppm for 5-ethyl substituted
compounds as compared to the 5-methyl counterparts.
A similar effect of the ethyl group on the shifts of C-4
and C-3 was observed: in compounds without C-4 sub-
stituent or with 4-methyl groups resonances were found
between 34 and 39 ppm, while the signals for trans- and
cis-4,5-EEMPO (with a 4-ethyl group) appeared at 44.2
and 47.8 ppm, respectively. For the spin traps without 3-
substituent or with 3-methyl substitution the C-3 signal
appeared at 32–34 ppm, whereas it was down-field
shifted to 40.4 and 40.2 ppm for trans-3,5-EEMPO
and cis-3,5-EEMPO, respectively. The shift of the
Ethyl groups generally showed a diastereotopic splitting
of the methylene protons, which was pronounced for
5-ethyl and 4-ethyl groups with two well-separated mul-
tiplets and less significant for 3-substituents with the sig-
nals of both protons partly overlapping. As an example,
in both cis- and trans-5,3-EEMPO, proton 5a-CH2A
appeared as a broad multiplet at 1.92–2.09 ppm and
5a-CH2B as a broad multiplet at 2.18–2.36 ppm.
The signal patterns of the H-3 and H-4 protons exhib-
ited a large dependence on substitution site and config-
uration, which is in complete agreement with our
previous work.1–3 In the 4-substituted derivatives the
resonance of H-4 gives a good indication of the config-
uration. The signal of trans-derivatives appeared at
significantly lower field than that of the cis-derivatives:
2.96 ppm versus 2.73 ppm for the 4-methyl derivative
and 2.72–2.81 ppm versus 2.22–2.27 ppm for 4-ethyl
compounds. The two geminal H-3 protons appeared at
about 2.30 ppm/2.90 ppm for the trans-compounds and
2.35/2.85 ppm for the cis-compounds, the shift differ-
ences between the two geminal ring protons being thus
slightly larger for the trans-isomers. In the trans-config-
ured 4-substituted derivatives, the high-field H-4 (about
2.30 ppm) is placed cis to the alkyl at C-4 and thus trans
to the ester. The down-field protons (about 2.90 ppm)
are consequently trans to the 4-alkyl and thus cis to
Figure 1. General structure of the spin traps.

K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
Table 1. ESI Q-TOF HR-MS analysis of the spin traps
2829
Sample
Acquired [MH]+ Calcd [MH]+ Error (ppm) mDa
Acquired [MNa]+ Calcd [MNa]+ Error (ppm) mDa
cis-3,5-EEMPO
trans-3,5-EEMPO
cis- 4,5-EEMPO
200.1321
200.1364
200.1343
200.1283
200.1283
200.1283
200.1283
200.1283
200.1283
200.1283
200.1283
18.89
40.67
30.13
69.06
30.38
17.69
52.77
25.08
3.78 222.1159
8.14 222.1204
6.03 222.1123
13.82 222.1134
6.08 222.1141
3.54 222.1122
10.56 222.1172
5.02 222.1189
222.1108
222.1108
222.1108
222.1108
222.1108
222.1108
222.1108
222.1108
23.05
43.40
6.75
5.12
9.64
1.50
2.58
3.28
1.44
6.40
8.10
trans- 4,5-EEMPO 200.1421
11.62
14.77
6.48
cis-5,3-EEMPO
trans-5,3-EEMPO
cis-5,4-EEMPO
trans-5,4-EEMPO
200.1344
200.1318
200.1389
200.1333
28.81
36.47
the ester. In the cis-4,5-EEMPO derivative, the situation
is opposite: the high-field proton (2.35 ppm) is now
configured trans to 4-alkyl, that is, cis to the 5-ethoxy-
carbonyl, and the down-field proton (2.85 ppm) cis to
4-alkyl, that is, trans to the 5-ethoxycarbonyl.
all spectra shown in Figure 2 were recorded 10 min after
mixing. In Figure 2a, the ESR spectrum of the cis-3,5-
EEMPO superoxide adduct is shown, consisting of
two different species. The major species (aN = 13.10 G;
aH = 7.81 G; 57%) has very broad lines, possibly indicat-
ing either the contribution of unresolved c-splittings or
the presence of two different stereoisomers (rotamers)
having only slightly different HFS parameters. The min-
or species (aN = 13.55 G; aH = 15.44 G; 43%) on the
other hand is characterized by very narrow lines. From
trans-3,5-EEMPO only one adduct was formed (see Ta-
ble 4). In contrast to 3,5-EEMPO, the respective 4,5-
EEMPO could not be separated into the cis- and
trans-forms using conventional chromatographic proce-
dures. When cis-/trans-4,5-EEMPO was incubated in the
hypoxanthine/xanthine oxidase system described above,
four different adducts were found in almost equal contri-
bution (Fig. 2b: aN = 13.40 G; aH = 9.55 G; 28% //
aN = 13.50 G; aH = 11.45 G; 27% // aN = 13.40 G;
aH = 6.45 G; 24% // and aN = 13.81 G; aH = 16.33 G;
21%). From trans-5,3-EEMPO (Fig. 2c) and cis-/trans-
5,4-EEMPO (Fig. 2d) two different adducts could be ob-
tained, which were, however, not completely resolved.
In 3-alkyl-substituted EEMPO derivatives, H-3 ap-
peared at around 3.04 ppm/3.17 ppm for the 3-methyl
derivative, and at 2.86 ppm/2.78 ppm for the 3-ethyl
counterpart. The shift difference between cis- and
trans-isomers was thus considerably smaller than that
of H-4 in 4-alkyl-substituted congeners. The splitting
of the geminal H-4 protons was very large for the
trans-compounds (1.62 ppm vs 2.57 ppm for 3-ethyl
and 1.76 ppm vs 2.62 ppm for 3-methyl). In cis-config-
ured 3-substituted derivatives, the splitting was signifi-
cantly smaller (2.11 ppm vs 2.20 ppm for 3-ethyl and
2.07 ppm vs 2.49 ppm for 3-methyl). In the trans-deriv-
atives, the high-field protons (1.72 and 1.76 ppm) are
configured cis to the alkyl groups at C-3 and C-5, and
trans to the 5-ethoxycarbonyl moiety, the down-field
protons (2.57 and 2.62 ppm) trans to both alkyl groups
and cis to the ester. In 3-substituted cis-compounds, the
situation is again reversed: the high-field protons (2.07
and 2.1 ppm) are now placed trans to the two vicinal al-
kyl groups at C-3 and C-5 (cis to the ester), the down-
field protons (2.20 and 2.49 ppm) are in cis-arrangement
to the alkyl groups (trans to the ester).
In order to determine the half-lives, samples were incu-
bated until maximum intensity was obtained. Then
SOD was added and the decay of the signal was re-
corded as a series of repetitive scans for at least
30 min. In this way the contribution of secondary spe-
cies could be subtracted from each individual spectrum
thus giving more accurate values for the half-lives of
the respective superoxide adducts. We also tried solid
KO2 (ca. 1 mg/ml, immediately followed by the addition
of SOD and catalase), which resulted in a higher signal-
to-noise ratio and a deviation from first order kinetics
during the first few minutes (half-life values given in
brackets). Despite this fact, the kinetic in the KO2 sys-
tem was in good agreement with a first order exponen-
tial decay (r2 > 0.95), which was used for calculations
of the half-lives (see Table 3). Furthermore, from all
3,5-EEMPO and 5,3-EEMPO isomers, separate half-
lives corresponding to the respective cis- and trans-forms
could be obtained. An overview of all relevant parame-
ters is given in Tables 3 and 4.
The 10-methylene groups in the ethoxy moieties of the
ester showed pronounced diastereotopic splitting. The
resonances were independent of the configuration, and
were found around 4.26 ppm for all derivatives with
the exception of 3,5-EEMPO (4.12 ppm). Also the
methyl resonance in the ethoxy group—generally about
1.30 ppm—was different for the 3,5-EEMPO derivatives
(1.17 ppm).
2.2. Spin trapping of superoxide radicals
We tested the spin trapping properties of the EMPO
derivatives toward superoxide radicals using an enzy-
matic system containing hypoxanthine (0.2 mM), xan-
thine oxidase (50 mU/ml), and the respective spin trap
(20 mM) in oxygenated phosphate buffer (20 mM), pH
7.4, containing 0.4 mM DTPA. Immediately after mix-
ing ESR spectra were recorded every 90 s, showing a
gradual increase in intensity until an optimal signal
intensity was reached about 10 min after mixing. At a la-
ter stage only a minor intensity increase was detected
and secondary products started to be formed. Therefore,
2.3. Spin trapping of hydroxyl radicals
For the generation of hydroxyl radicals we used a Fen-
ton-type system6 consisting of the spin trap (40 mM),
hydrogen peroxide (0.2%), EDTA (2 mM), and iron-
(II) sulfate (1 mM) in water. After 10 s, the reaction

2830
K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
Figure 2. Formation of the superoxide adducts of the spin traps cis-
3,5-EEMPO, cis-/trans-4,5-EEMPO, trans-5,3-EEMPO, and cis-/trans-
5,4-EEMPO. (a) cis-3,5-EEMPO (20 mM), catalase (250 U/mL),
hypoxanthine (0.2 mM), and xanthine oxidase (50 mU/mL) in oxy-
genated phosphate buffer (20 mM, pH 7.4, containing 0.4 mM DTPA)
were incubated and measured using the following EPR parameters:
sweep width, 60 G; modulation amplitude, 0.52 G; microwave power,
20 mW; time constant, 0.08 s; receiver gain, 1 · 105; scan rate, 42.9 G/
min. (b) Same as in (a), except that cis-/trans-4,5-EEMPO (20 mM)
was used. (c) Same as in (a), except that trans-5,3-EEMPO (20 mM)
was used; sweep width, 80 G; modulation amplitude, 1.05 G; micro-
wave power, 20 mW; time constant, 0.08 s; receiver gain, 5 · 104; scan
rate, 57.3 G/min. (d) Same as in (a), except that cis-/trans-5,4-EEMPO
(20 mM) was used. The bars represent 5000 arbitrary units.
was stopped upon 1:1 dilution with DTPA (20 mM) in
phosphate buffer (300 mM, pH 7.4). In Figure 3a, the
hydroxyl radical adduct of cis-3,5-EEMPO is shown,
consisting of three different species (aN = 14.59 G;
aH = 19.34; 70%, aN = 13.89 G; aH = 7.94; 20%, and
aN = 13.92 G; aH = 6.18 G, 10%). The respective data
obtained from the hydroxyl radical adducts of cis-/
trans-4,5-EEMPO
(Fig. 3c), and cis-/trans-5,4-EEMPO (Fig. 3d) are listed
(Fig.
3b),
trans-5,3-EEMPO
in Table 4.
2.4. Spin adducts formed from methanol: methoxyl and
hydroxymethyl radical adducts
Methanol was chosen as a model compound for alco-
hols, from which two different types of radical adducts
can be formed: first, the oxygen-centered methoxyl rad-
ical adducts (Fig. 4), and second, the carbon-centered
hydroxymethyl radical adducts (Fig. 5).
The methoxyl radical adducts were synthesized by nucle-
ophilic addition of methanol using a model system pre-
viously described by Dikalov and Mason7 and already
tested with other spin traps.8–11 In Figure 4a, trans-

K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
2831
Table 3. Half-life of the superoxide adducts and n-octanol/buffer
partition coefficients of the spin traps
were obtained with cis-/trans-5,4-EEMPO and peroxi-
dized linoleic acid (Fig. 6c, aN = 12.05 G; aH = 7.30 G;
64%, and aN = 12.05 G; aH = 10.30 G; 36%) or tert-bu-
tyl hydroperoxide (Fig. 6d, aN = 12.15 G; aH = 7.61 G;
50%, and aN = 12.15 G; aH = 10.65 G; 50%). From the
other spin traps no alkoxyl radical adducts were ob-
tained. Instead, weak and rather complicated spectra
were detected which could, however, not be identified.
Compound
Apparent
t1/2(min)
XOD (KO2)
Partition coefficient
n-octanol/phosphate
buffer (100 mM, pH 7.0)
EMPO*
8.6*
0.15*
cis-3,5-EEMPO 12.32 (10.54) 1.38
trans-3,5-EEMPO 45.09 (33.98) 1.67
cis/trans-4,5-EEMPO 37.07 (33.90) 1.34
cis-5,3-EEMPO
trans-5,3-EEMPO
cis/trans-5,4-EEMPO 28.09 (34.72) 1.07
—
55.04 (30.75) 1.48
(22.71) 1.34
3. Discussion
* Data from Stolze et al.9
Eight novel EMPO derivatives (four of them as pure cis-
or trans-forms) were synthesized in this study, bearing
ethyl substituents in positions 3, 4, or 5, the remaining
substituents being a methyl group and a 5-ethoxycar-
bonyl substituent in all cases. The structure of the com-
pounds was comprehensively characterized by full
NMR assignment (1H and 13C), ESI Q-TOF HR-MS,
and IR spectroscopy. Determination of the half-lives
of the superoxide adducts was done using a first-order
exponential decay approximation (Table 3, Pearson cor-
relation coefficient r2 > 0.95). Using a KO2 system (high-
er superoxide formation rate) a contribution of second
order decay of the superoxide adducts was observed
during the first few minutes.
3,5-EEMPO/ÅOCH3 was obtained in a methanolic iron-
(III) chloride solution which after 2 min was diluted 1:50
with 300 mM phosphate buffer containing 20 mM
DTPA.
cis-/trans-4,5-EEMPO/ÅOCH3 was obtained in the same
way (Fig. 4b), but only the trans-5,3-EEMPO/ÅCH2OH
adduct could be detected (Fig. 4c) while the expected
trans-5,3-EEMPO/ÅOCH3 seemed to be unstable. The
cis-/trans-5,4-EEMPO/ÅOCH3 adduct is shown in Figure
4d.
For the formation of the hydroxymethyl radical ad-
ducts a Fenton-type system6 was chosen consisting of
the spin trap (40 mM), hydrogen peroxide (0.2%),
EDTA (2 mM), and iron-(II) sulfate (1 mM) in metha-
nol/water (20/80, v/v). After 10 s, the reaction was
stopped upon 1:1 dilution with DTPA (20 mM) in
phosphate buffer (300 mM, pH 7.4). In Figure 5a,
the hydroxymethyl radical adduct of cis-3,5-EEMPO
is shown, consisting of two different species
In addition to the half-life of the spin adducts the rate
constant of the spin trapping reaction, the spectral line
width, the total number of lines, and additional factors,
such as enzyme binding or the solubility (aggregate for-
mation) of the spin trap, reduction by antioxidants or
buffer constituents, and degradation by transition metal
catalyzed reactions (Fenton type reactions, oxidation by
Fe3+, etc.) play also important roles regarding the over-
all efficiency of the spin trapping reaction.
(aN = 15.00 G; aH = 24.85 G;
a
H(2) = 0.50 G; aH(3)
=
=
0.45 G; 77%, and aN = 14.97 G; aH = 14.00 G; aH(3)
0.56 G; aH(2) = 0.30 G; 23%), which can clearly be dis-
tinguished from the methoxyl radical adduct shown
above (in Fig. 4a). The respective data obtained from
the hydroxymethyl radical adducts of cis-/trans-4,5-
EEMPO (Fig. 5b), trans-5,3-EEMPO (Fig. 5c), and
cis-/trans-5,4-EEMPO (Fig. 5d) are listed in Table 4.
In view of these facts the half-life values given in Table 3
are therefore to be interpreted only as a preliminary
estimation of the spin trapping performance.
Hydroxyl radical adducts were stable for more than
20 min, whereas the respective methoxyl radical adducts
were rather unstable and disappeared within the first 5–
10 min (3,5-EEMPO, 4,5-EEMPO, and 5,4-EEMPO) or
were not detected at all (from 5,3-EEMPO). From perox-
idized linoleic acid lipid-derived radical adducts could be
detected in toluene, whereas in aqueous solution the
respective hydroxyl radical adducts were the predominant
species. Alkoxyl radicals undergo rapid b-scission (rate
constant ca. 106 Mꢀ1 sꢀ1) leading to a series of rearranged
products14–20 and are therefore difficult to detect. On the
other hand, UV-irradiation of peroxides in toluene could
be used for the detection of both, alkoxyl as well as per-
oxyl radical adducts,13 depending on the ratio applied
between the spin trap and the respective hydroperoxide.
2.5. Spin trapping of lipid-derived radicals
Trying to detect lipid-derived free radicals using a re-
cently described Fenton type system8,12 containing
LOOH instead of hydrogen peroxide mainly resulted
in the detection of the hydroxyl radical adduct. We
therefore decided to use a UV-irradiated system in tolu-
13
´
ene previously reported by Clement et al.
In Figure 6a, the radical adduct obtained from cis-3,5-
EEMPO (50 mM) and excess peroxidized linoleic acid
(ca. 100 mM) is shown, consisting of one major species
(aN = 12.25 G; aH = 14.10 G). In Figure 6b, the experi-
ment was performed with excess tert-butyl hydroperox-
ide (ca. 300 mM), the spectrum most probably showing
the tert-butoxyl radical adduct (aN = 12.14 G;
aH = 13.06 G; aH = 0.62 G) which decayed rapidly,
thereby forming a complicated mixture of secondary
products which could not be identified. Similar results
4. Conclusion
In conclusion, four of the eight possible diastereomeric
forms of the investigated EEMPO compounds were
successfully separated and can be recommended for

Table 4. Comparison of the EPR parameters of different radical adducts of various EEMPO compounds
Radical
ÅOOH
HFS (G)
EMPO*
c-3,5-EEMPO
t-3,5-EEMPO
4,5-EEMPO
c-5,3-EEMPO
t-5,3-EEMPO
5,4-EEMPO
t-3,5-EDPO*
4,5-EDPO*
(57%) (43%)
(57%) (43%) (100%)
28%/27%/
24%/21%
(50%) (50%) (77%) (23%)
13.03 13.03 13.03 13.00
(57%) (43%)
(100%)
(36%) (35%) (29%)
aN
aH
aH
13.28 13.28
11.89 9.67
13.10 13.55 13.10
13.40/13.50/
13.40/13.81
9.55/11.45/
6.45/16.33
13.45 13.46
10.80 8.18
13.55
5.98
—
13.48 13.48 13.48
7.81
—
15.44 7.89
8.26
—
5.86
—
7.21
—
8.79
—
9.10
—
11.60 8.55
— 1.90
—
—
—
—
—
—
—
—
—
—
ÅOH
(50%) (50%) (70%) (20%) (10%) (84%) (16%) (71%) (29%) (51%) (49%) (66%) (34%) (64%) (23%) (13%)
14.00 14.00 14.59 13.89 13.92 13.80 14.26 13.97 14.51 13.80 14.29 13.66 14.11 13.80 13.84 14.17
(67%) (33%)
13.85 14.30
(75%) (25%)
14.00 14.45
10.00 18.60
aN
aH
aH
aH
15.00 12.58 19.34 7.94
6.18
—
8.57
0.83
0.344
18.80 9.04
19.12 9.24
19.28 8.52
—
19.45 10.81 8.38
19.90
—
9.00
0.75
19.30
0.95
0.90
—
—
—
—
—
0.93
0.344
1.21
0.58
—
—
—
—
—
1.05
—
—
(100%)
15.52
22.21
20.82
(100%)
15.58
26.16
16.01
(0.56)
(100%)
15.47
25.76
17.72
(0.54)
(100%)
15.63
26.25
16.73
0.57
(100%)
15.37
24.88
17.08
(100%)
(50%) (50%)
15.37 15.12
26.06 24.14
17.60 18.70
(100%)
15.45
25.87
17.50
(100%)
15.57
25.31
18.03
ÅH
aN
aH
aH
15.23
26.40
16.57
ÅCH3
(100%)
15.42
22.30
—
(60%) (40%) (90%) (10%) (100%)
15.32 15.45 15.32 15.33 15.43
14.38 26.45 15.23 25.59 17.03
(52%) (48%) (82%) (18%)
15.22 15.12 15.02 15.11
26.56 18.67 15.68 26.19
(54%) (46%)
14.67 14.80
15.94 22.58
(89%) (11%)
15.32 15.35
16.74 26.20
(98%) (2%)
15.40 15.50
18.43 26.73
aN
aH
aH
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
ÅOCH3
(50%) (50%)
13.74 13.74
10.87 7.81
(62%) (38%) (67%) (33%) (57%) (43%) (56%) (44%) (56%) (44%)
13.61 14.05 13.55 14.00 13.61 13.61 13.41 13.73 13.34 13.86
(56%) (44%)
13.51 13.51
(61%) (39%) (56%) (24%) (20%)
13.55 13.95 13.57 13.52 14.13
aN
aH
aH
4.25
1.00
16.03 6.26
1.05 1.05
15.75 8.11
1.00 0.87
4.09
0.70
4.70
0.88
14.80 6.39
— 0.95
16.88
1.05
5.62
0.60
8.88
0.60
6.77
0.99
16.10 8.55
0.95 0.70
4.50
—
16.43
—
—
—
ÅCH2OH
(100%)
14.95
21.25
—
(77%) (23%) (59%) (41%) (71%) (29%) (56%) (44%) (58%) (42%) (100%)
15.00 14.97 14.76 14.95 15.06 14.97 14.90 14.70 14.74 14.63 14.74
24.85 14.00 16.42 23.83 16.87 18.32 24.83 16.76 24.37 16.74 18.62
(64%) (36%)
14.73 14.90
17.10 24.33
(98%) (2%)
15.00 15.00
18.20 25.30
aN
aH
aH
aH
0.453
0.502
0.563
0.302
0.453
0.502
0.453
0.63
0.44
0.502
0.342
0.502
0.502
0.502
0.502
(0.50)2
—
—
ÅCH(OH)CH3
(67%) (33%) (100%)
14.94 15.00 14.80
20.82 22.40 22.76
(58%) (42%) (63%) (37%) (77%) (23%) (83%) (17%) (100%)**
14.80 14.69 15.07 15.05 14.91 14.40 14.66 14.67 14.74
22.18 16.08 17.63 19.24 23.62 18.96 23.32 16.46 19.12
(63%) (37%) (100%)
14.84 14.64 15.10
23.30 17.07 18.46
aN
aH
ꢀ
ÅCO2
(100%)
14.74
17.16
—
(100%)
14.99
14.99
—
(100%)
14.84
22.74
—
(80%) (20%) (50%) (50%) (50%) (50%)
15.30 14.90 14.80 14.67 14.80 14.67
15.30 23.68 22.63 16.02 22.63 16.02
(90%) (10%)
14.87 12.39
16.78 12.39
(97%) (3%)
14.84 14.84
15.55 22.80
(95%) (5%)
14.90 14.78
15.80 22.70
aN
aH
aH
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
1.48
0.384
0.304
0.354
0.304
(LOOH)
(62%) (38%) (100%)**
13.45 13.45 12.25
—
(64%) (36%)** (50%) (45%)
(68%) (32%)
13.85 14.60
aN
aH
aH
—
12.05 12.05
14.40 13.75
18.90 8.83
11.45 8.55
14.10
—
—
7.30
10.30
—
9.90
1.05
18.30
—
—
—
—
—
—
—
—
(tBuOOH)
(100%)**
12.14
13.06
0.62
(50%) (50%)**
12.15 12.15
aN
aH
aH
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
7.61
—
10.65
—
—
* Data from Stolze et al.2
** High line width: unresolved lines.

K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
2833
Figure 3. Formation of hydroxyl radical adducts of the spin traps cis-
3,5-EEMPO, cis-/trans-4,5-EEMPO, trans-5,3-EEMPO, and cis-/trans-
5,4-EEMPO. (a) cis-3,5-EEMPO (40 mM) was incubated with a
Fenton system containing FeSO4 (1 mM), EDTA (2 mM), H2O2
(0.2%). The reaction was stopped after 10 s by 1:1 dilution with
phosphate buffer (300 mM, pH 7.4, containing 20 mM DTPA) and the
spectrum was recorded using the following spectrometer settings:
sweep width, 80 G; modulation amplitude, 0.23; microwave power,
20 mW; time constant, 0.08 s; receiver gain, 1 · 104; scan rate, 57.2 G/
min. (b) Same as in (a), except that cis-/trans-4,5-EEMPO (40 mM)
was used. (c) Same as in (a), except that trans-5,3-EEMPO (40 mM)
was used. (d) Same as in (a), except that cis-/trans-5,4-EEMPO
(40 mM) was used. The bars represent 5000 arbitrary units.
Figure 4. Iron catalyzed addition of methanol to the spin traps cis-3,5-
EEMPO, cis-/trans-4,5-EEMPO, trans-5,3-EEMPO, and cis-/trans-5,4-
EEMPO. (a) After 2 min incubation of cis-3,5-EEMPO (1 M in
methanol) with FeCl3 (2 mM), the reaction was stopped by 1:20
dilution with phosphate buffer (0.15 M, pH 7.4, containing 10 mM
DTPA), and the spectrum was recorded with the following spectrom-
eter settings: sweep width, 80 G; modulation amplitude, 0.23 G;
microwave power, 20 mW; time constant, 0.08 s; receiver gain,
1 · 104; scan rate, 57.2 G/min. (b) Same as in (a), except that cis-/
trans-4,5-EEMPO was used. (c) Same as in (a), except that trans-5,3-
EEMPO was used. (d) Same as in (a), except that cis-/trans-5,4-
EEMPO was used. The bars represent 1000 arbitrary units.
trapping of superoxide radicals when higher lipophilic
properties are required. The half-lives of their superox-
ide adducts are greatly dependent on the stereochemical
structure, the highest values being comparable to the
respective values observed with the EDPO compounds
presented earlier.2 Toxicological properties are presently
being tested and will be published in the near future.
There is no correlation between lipophilicity and the
half-life times, but hydrophilic compounds such as
DMPO, DEPMPO, EMPO, or iPrMPO seem to be less
toxic and therefore better suitable for biological studies
than the more lipophilic spin traps PBN and 5-t-butoxy-
carbonyl-5-methyl-1-pyrroline N-oxide.
5.2. Syntheses
Synthesis and characterization of the compounds were
performed as reported previously,1–3 in analogy to the
synthesis of EMPO and its derivatives21–23 with minor
adaptations as given below.
5.2.1. Ethyl 2-bromobutanoate and ethyl 2-bromo-pro-
panoate. 2-Bromobutanoyl bromide (or 2-Bromopropio-
nyl bromide) (70 mmol) was slowly added to a solution
of ethanol (100 mmol) and pyridine (70 mmol) in CHCl3
at 0 ꢁC (ice bath). After stirring for 1 h, the reaction
mixture was successively washed with water (50 ml), sul-
furic acid (10%, 50 ml), and concentrated aqueous
NaHCO3 (50 ml), and dried over Na2SO4 overnight.
Solvent and excess ethanol were removed under reduced
pressure. The crude, nearly colorless product was used
without further purification.
5. Experimental
5.1. Chemicals
2-Bromobutanoyl bromide, 2-bromopropionyl bromide,
crotonic aldehyde, methacrolein, and trans-2-pentenal
were purchased from Sigma–Aldrich. Petroleum ether
(high boiling, 50–70 ꢁC) was obtained from VWR
BDH Prolabo and distilled twice before use. All other
chemicals were from Merck.
5.2.2. Ethyl 2-nitrobutanoate (or ethyl 2-nitropropano-
ate). The respective ethyl 2-bromoalkanoate (60 mmol)
was added under stirring to a solution of sodium nitrite
(7.2 g, 104 mmol) and phloroglucinol dihydrate (8.5 g,
52 mmol) in dry N,N-dimethylformamide (120 ml) at

2834
K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
Figure 5. Formation of hydroxymethyl radical adducts of the spin
traps cis-3,5-EEMPO, cis-/trans-4,5-EEMPO, trans-5,3-EEMPO, and
cis-/trans-5,4-EEMPO. (a) cis-3,5-EEMPO (40 mM) was incubated
with Fenton system containing FeSO4 (1 mM), EDTA (2 mM), H2O2
(0.2%) in the presence of 10% methanol. The reaction was stopped
after 10 s by 1:1 dilution with phosphate buffer (300 mM, pH 7.4,
containing 20 mM DTPA) and the spectrum was recorded using the
following spectrometer settings: sweep width, 80 G; modulation
amplitude, 0.23 G; microwave power, 20 mW; time constant, 0.08 s;
receiver gain, 1 · 104; scan rate, 57.2 G/min. (b) Same as in (a), except
that cis-/trans-4,5-EEMPO (40 mM) was used. (c) Same as in (a),
except that trans-5,3-EEMPO (40 mM) was used. (d) Same as in (a),
except that cis-/trans-5,4-EEMPO (40 mM) was used. The bars
represent 5000 arbitrary units.
Figure 6. Hydroperoxide-derived free radical adducts of cis-3,5-
EEMPO and cis-/trans-5,4-EEMPO formed from peroxidized linoleic
acid and tert-butylhydroperoxide in UV-irradiated toluene solution
(a)
a nitrogen-bubbled solution of peroxidized linoleic acid
(100 mM) and cis-3,5-EEMPO (50 mM) in toluene was irradiated
for 10 s and the spectrum was recorded with the following
spectrometer settings: sweep width, 70 G; modulation amplitude,
0.21 G; microwave power, 10 mW; time constant, 0.04 s; receiver
gain, 1 · 105; scan rate, 100.2 G/min. (a)
A nitrogen-bubbled
solution of tert-butylhydroperoxide (300 mM) and cis-3,5-EEMPO
(50 mM) in toluene was irradiated for 10 s and the spectrum was
recorded with the following spectrometer settings: sweep width,
70 G; modulation amplitude, 0.1 G; microwave power, 10 mW; time
constant, 0.08 s; receiver gain, 1 · 104; scan rate, 50.1 G/min. (c)
Same as in (a), except that cis-/trans-5,4-EEMPO was used.
Spectrometer settings: sweep width, 70 G; modulation amplitude,
0.21 G; microwave power, 10 mW; time constant, 0.02 s; receiver
gain, 1 · 105; scan rate, 200.3 G/min. (d) Same as in (b), except that
cis-/trans-5,4-EEMPO was used. Spectrometer settings: sweep width,
70 G; modulation amplitude, 0.1 G; microwave power, 10 mW; time
constant, 0.08 s; receiver gain, 1 · 104; scan rate, 50.1 G/min. The
bars represent 5000 arbitrary units.
room temperature. The solution was stirred overnight,
poured into ice water (240 ml), and extracted 4 times with
ethyl acetate (100 ml). The combined extracts were trea-
ted twice with 100 ml of saturated NaHCO3 solution
and dried over Na2SO4. After removal of the solids by
filtration, the solvent was evaporated in vacuo. The ob-
tained colorless or pale yellow products were used without
further purification.
5.2.3. Ethyl 2-ethyl-4-formyl-2-nitropentanoate. Ethyl 2-
nitrobutanoate (3.7 g, 23 mmol) was dissolved in a mix-
ture of acetonitrile (10 g, 244 mmol) and triethylamine
(0.2 g, 2 mmol). Methacrolein (2.66 g, 38 mmol) was
slowly added at 0 ꢁC. The solution was kept at 10 ꢁC
overnight and then poured into a solution of ice-cold
HCl (5 ml of concentrated HCl in 150 ml of water).
The solution was extracted 3 times with CH2Cl2 and
dried over Na2SO4. After filtration, the mixture was dis-
tilled under reduced pressure, and the purity of the
remaining product was assessed by thin layer chroma-
tography and IR spectroscopy (Table 2).
(2.66 g, 38 mmol) was used instead of methacrolein. In
addition, the solution had to be stirred for several days
at room temperature.
5.2.5. Ethyl 4-formyl-2-methyl-2-nitrohexanoate. Same
conditions as above, except that ethyl 2-nitropropanoate
(3.4 g, 23 mmol) and ethacrolein (3.2 g, 38 mmol) were
used instead of ethyl 2-nitrobutanoate and methacro-
lein, respectively.
5.2.6. Ethyl 3-ethyl-4-formyl-2-methyl-2-nitrobutanoate.
Same conditions as above, except that ethyl 2-nitropro-
panoate (3.4 g, 23 mmol) and trans-2-pentenal (3.2 g,
38 mmol) were used instead of ethyl 2-nitrobutanoate
and methacrolein, respectively.
5.2.4. Ethyl 2-ethyl-4-formyl-3-methyl-2-nitro-butanoate.
Same conditions as above, except that crotonic aldehyde

K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
2835
5.2.7. Synthesis of the N-oxides. Synthesis of the nitrones
was performed according to the procedure described re-
cently for the synthesis of EMPO derivatives.1–3,21 To a
concentrated solution of 25 mmol of the respective ethyl
2-alkyl-4-formyl-2-nitroalkanoate in H2O/CH3OH (v/
v = 6:4), an aqueous solution of ammonium chloride
(1.87 g in 8 ml of water) was added. While zinc dust
(8.5 g, 130 mmol) was slowly added within 30 min, the
mixture was carefully kept at room temperature. The
mixture was stirred for 4.5 h at room temperature, the
white precipitate and the remaining zinc powder were re-
moved by filtration, and the residue was washed five
times with methanol (30 ml). The liquid phase was con-
centrated to about 10 ml, saturated with borax, and ex-
tracted 4 times with 60 ml CH2Cl2. The organic phase
was dried with Na2SO4, filtered, and concentrated.
Careful purification by column chromatography on sil-
ica gel with a petroleum ether/ethanol gradient allowed
the separation from the majority of side products and
provided the product as a yellow oil or yellow needles.
Additional purification was done immediately before
the EPR experiments on a 1 ml solid phase extraction
column using a Chromabond C-18 100 mg column ob-
5.2.8.3. trans-5,3-EEMPO. 1H NMR: 00.97 (t, 3H,
5bCH3), 1.20 (d, 3H, 3aCH3), 1.29 (t, 3H, 2 CH3), 1.76
(dd, 1H, CHB, trans to ester, cis to Et and Me), 1.92–
4
2.09 (m, 1H, 5aCH2A), 2.18–2.36 (m, 1H, 5aCH2B), 2.62
4
(dd, 1H, CHA, cis to ester, trans0 to Et and Me), 3.17
(m, 1H, 3CH), 4.26 (m, 2H, 1 CH2), 6.91 (d, 1H,
3J = 3.6 Hz, 2CH). 13C NMR: 7.2 (5bCH3), 13.9
0
(2 CH3), 18.4 (3aCH3), 25.8 (5aCH2), 33.7 (3CH), 36.4
0
(4CH2), 62.0 (OA1 CH2); 83.1 (5C); 140.4 (N@2CH),
170.1 (COO).
5.2.8.4. cis-5,3-EEMPO. 1H NMR: 0.95 (t, 3H,
0
5bCH3), 1.25 (d, 3H, 3aCH3), 1.31 (t, 3H, 2 CH3), 1.92–
4
2.09 (m, 1H, 5aCH2A), 2.07 (m, 1H, CHB, cis to ester),
2.18–2.36 (m, 1H, 5aCH2B), 2.49 (dd, 1H, 04CHA, trans to
3
ester), 3.04 (m, 1H, CH), 4.26 (m, 2H, 1 CH2), 6.90 (d,
1H, 3J = 2.7 Hz, 2CH). 13C NMR: 7.0 (5bCH3), 13.8
0
(2 CH3), 19.1 (3aCH3), 25.2 (5aCH2), 33.1 (3CH), 36.1
0
(4CH2), 62.0 (OA1 CH2); 83.0 (5C); 139.8 (N@2CH),
169.8 (COO).
5.2.8.5. trans-4,5-EEMPO. 1H NMR: 0.93 (d, 3H,
0
4bCH3), 1.31 (t, 3H, 2 CH3), 1.34–1.45 (m, 1H, 4aCH2A),
1.55–1.66 (m, 1H, 4aCH2B), 1.61 (s, 3H, 5aCH3), 2.30–
tained from Macherey-Nagel (Duren, Germany). The
¨
purity of the obtained products was assessed by TLC
and UV spectroscopy. Final identification of the purified
3
2.48 (m, 1H, CH2B), 2.72–2.81 (m, 1H, 4CH), 2.83–
0
3
2.94 (m, 1H, CH2A), 4.27 (m, 2H, 1 CH2), 7.05 (t, 1H,
1
products was performed by H NMR, 13C NMR, ESI
3J = 2.3 Hz, 2CH). 13C NMR: 11.6 (4bCH3), 13.9
0
Q-TOF HR-MS (Table 1), and IR spectroscopy (Table
2).
(2 CH3), 14.8 (5aCH3), 22.7 (4aCH2), 32.7 (3CH2), 44.2
0
(4CH), 62.3 (OA1 CH2); 82.1 (5C); 136.1 (N@2CH),
170.0 (COO).
5.2.8. NMR. 1H NMR spectra were recorded at
300.13 MHz, 13C NMR spectra at 75.47 MHz with
CDCl3 as the solvent and TMS as the internal stan-
dard - if not stated otherwise. Data are given in
ppm units. 13C peaks were assigned by means of
HMQC and HMBC spectra. The numbering of car-
bons follows the conventions in spin trap studies, it
should be noted that the numbering of the heterocyclic
atoms runs opposite to conventional heterocycle
nomenclature (C-2 is the double-bonded carbon, while
C-5 is the quaternary C carrying the ethoxycarbonyl
group).
5.2.8.6. cis-4,5-EEMPO. 1H NMR: 0.97 (d, 3H,
0
4bCH3), 1.31 (t, 3H, 2 CH3), 1.34–1.45 (m, 1H, 4aCH2A),
1.55–1.66 (m, 1H, 4aCH2B), 1.70 (s, 3H, 5aCH3), 2.22–
4
2.27 (m, 1H, CH), 2.32–2.43 (m, 1H, 3CH2B), 2.80–
0
3
2.90 (m, 1H, CH2A), 4.27 (m, 2H, 1 CH2), 7.22 (t, 1H,
3J = 2.6 Hz, 2CH). 13C NMR: 11.9 (4bCH3), 14.1
0
(2 CH3), 20.5 (5aCH3), 23.3 (4aCH2), 32.4 (3CH2), 47.8
0
(4CH), 62.1 (OA1 CH2); 81.9 (5C); 138.3 (N@2CH),
168.1 (COO).
5.2.8.7. trans-5,4-EEMPO. 1H NMR: 1.02 (d, 3H,
0
5bCH3), 1.18 (d, 3H, 4aCH3), 1.31 (t, 3H, 2 bCH3),
1.91–2.11 (m, 1H, 5aCH2A), 2.13–2.28 (m, 1H,
1
5.2.8.1. trans-3,5-EEMPO. H NMR: 0.85 (t, 3H,
0
3bCH3), 1.17 (m, 3H, 2 CH3), 1.28–1.54 (m, 2H,
3aCH2), 1.59 (s, 3H, 5aCH3), 1.62 (dd, 1H, 4CHB,
trans to ester, cis to Et and Me), 2.57 (dd, 1H,
5aCH2B), 2.32 (m, 1H, CH2B), 2.91 (m, 1H, CH2A),
3
3
0
2.96 (m, 1H, 4CH), 4.27 (m, 2H, 1 CH2), 7.10 (t,
1H, 3J = 2.3 Hz, 2CH). 13C NMR: 8.6 (5bCH3), 13.9
0
4CHA, cis to ester, trans to Et and Me), 2.86 (m,
(2 CH3), 14.3 (4aCH3), 22.6 (5aCH2), 35.1 (3CH2),
0
0
1H, 3CH), 4.12 (m, 2H, 1 CH2), 6.82 (d, 1H,
37.0 (4CH), 61.9 (OA1 CH2); 84.8 (5C); 136.0
(N@2CH), 169.7 (COO).
3J = 2.0 Hz, 2CH). 13C NMR: 11.4 (3bCH3), 13.9
0
(2 CH3), 21.7 (5aCH3), 26.5 (3aCH2), 39.2 (4CH2),
0
40.4 (3CH), 62.2 (OA1 CH2); 79.4 (5C); 139.0
(N@2CH), 170.1 (COO).
5.2.8.8. cis-5,4-EEMPO. 0.980 (d, 3H, 5bCH3), 1.09
(d, 3H, 4aCH3), 1.31 (t, 3H, 2 bCH3), 1.91–2.11 (m,
1H, 5aCH2A), 2.13–2.28 (m, 1H, 5aCH2B), 2.41 (m,
1H, 3CH2B), 2.73 (m, 1H, 0 4CH), 2.81 (m, 1H,
3CH2A), 4.27 (m, 2H, 1 CH2), 7.18 (t, 1H,
5.2.8.2. cis-3,5-EEMPO. 1H NMR: 0.85 (t, 3H,
0
3bCH3), 1.17 (m, 3H, 2 CH3), 1.28–1.54 (m, 2H,
3aCH2), 1.58 (s, 3H, 5aCH3), 2.11 (m, 1H, 4CHB, cis
3J = 2.3 Hz, 2CH). 13C NMR: 7.0 (5bCH3), 14.0
0
to ester), 2.20 (m, 1H, 4CHB, trans to ester), 2.78
(2 CH3), 15.2 (4aCH3), 23.07 (5aCH2), 34.21 (3CH2),
34.25 (4CH), 61.8 (OA1 CH2); 85.4 (5C); 138.0
(N@2CH), 168.2 (COO).
0
(m, 1H, 3CH), 4.12 (m, 2H, 1 CH2), 6.78 (d, 1H,
3J = 2.2 Hz, 2CH). 13C NMR: 11.6 (3bCH3), 14.0
0
(2 CH3), 21.4 (5aCH3), 26.4 (3a CH2), 38.1 (4CH2),
0
40.2 (3CH), 62.2 (OA1 CH2); 79.4 (5C); 138.0
(N@2CH), 169.9 (COO).
5.2.9. Preparation of lipid hydroperoxides. Linoleic acid
hydroperoxide was synthesized according to O’Brien.24

2836
K. Stolze et al. / Bioorg. Med. Chem. 15 (2007) 2827–2836
Briefly, linoleic acid was air-oxidized for 72 h in the
dark at room temperature. The oxidation mixture
was dissolved in petroleum ether (boiling range
60–90 ꢁC) and extracted four times with water/metha-
nol (v/v = 1:3). The obtained methanolic phase was
counter-extracted four times with petroleum ether
(boiling range 60–90 ꢁC) and evaporated under re-
duced pressure. The obtained hydroperoxide was dis-
solved in ethanol and stored in liquid nitrogen. The
concentration of hydroperoxide was determined by
UV spectroscopy based on an extinction coefficient
of e233nm = 25250 Mꢀ1 cmꢀ1 in ethanol.24
Acknowledgment
The authors wish to thank P. Jodl for skilful technical
assistance in synthesis, purification, and characteriza-
tion of the spin traps.
References and notes
1. Stolze, K.; Udilova, N.; Rosenau, T.; Hofinger, A.; Nohl,
H. Biochem. Pharmacol. 2005, 69, 297.
2. Stolze, K.; Rohr-Udilova, N.; Rosenau, T.; Stadtmuller,
¨
R.; Nohl, H. Biochem. Pharmacol. 2005, 69, 1351.
3. Stolze, K.; Rohr-Udilova, N.; Rosenau, T.; Hofinger, A.;
Kolarich, D.; Nohl, H. Bioorg. Med. Chem. 2006, 14, 3368.
4. Tsai, P.; Ichikawa, K.; Mailer, C.; Pou, S.; Halpern, H. J.;
Robinson, B. H.; Nielsen, R.; Rosen, G. M. J. Org. Chem.
2003, 68, 7811.
5. Tsai, P.; Marra, J. M.; Pou, S.; Bowman, M. K.; Rosen,
G. M. J. Org. Chem. 2005, 70, 7093.
6. Roubaud, V.; Lauricella, R.; Bouteiller, J. C.; Tuccio, B.
Arch. Biochem. Biophys. 2002, 397, 51.
5.3. Instruments
UV–vis spectra were recorded between 200 and 350 nm
on Hitachi 150–20 and U-3300 spectrophotometers in
double-beam mode. IR spectra were recorded as film
on an ATI Mattson Genesis Series FTIR spectrometer
(see also Table 2).
7. Dikalov, S. I.; Mason, R. P. Free Radical Biol. Med. 2001,
30, 187.
8. Stolze, K.; Udilova, N.; Nohl, H. Free Radical Biol. Med.
2000, 29, 1005.
9. Stolze, K.;Udilova, N.;Nohl, H. Biol. Chem. 2002, 383, 813.
10. Stolze, K.; Udilova, N.; Nohl, H. Biochem. Pharmacol.
2002, 63, 1465.
For EPR experiments, Bruker spectrometers (ESP300E
and EMX) were used, operating at 9.7 GHz with
100 kHz modulation frequency, equipped with a rectan-
gular TE102 or a TM110 microwave cavity.
NMR spectra were recorded on a Bruker Avance at
300.13 MHz for 1H, and 75.47 MHz for 13C. CDCl3
was used as the solvent throughout, TMS (tetra-
methylsilane) as the internal standard. 13C peaks
were assigned by means of APT (attached proton
test), HMQC (1H-detected heteronuclear multiple-
quantum coherence), and HMBC (heteronuclear mul-
tiple bond connectivity) spectra. All chemical shift
data are given in ppm units, coupling constants in
Hz.
11. Stolze, K.; Udilova, N.; Rosenau, T.; Hofinger, A.; Nohl,
H. Biol. Chem. 2003, 384, 493.
12. Stolze, K.; Udilova, N.; Nohl, H. Acta Biochim. Pol. 2000,
47, 923.
´
´
13. Clement, J.-L.; Ferre, N.; Siri, D.; Karoui, H.; Rocken-
bauer, A.; Tordo, P. J. Org. Chem. 2005, 70, 1198.
14. Rota, C.; Barr, D. P.; Martin, M. V.; Guengerich, F. P.;
Tomasi, A.; Mason, R. P. Biochem. J. 1997, 328, 565.
15. Van der Zee, J.; Barr, D. P.; Mason, R. P. Free Radical
Biol. Med. 1996, 20, 199.
16. Akaike, T.; Sato, K.; Ijiri, S.; Miyamoto, Y.; Kohno, M.;
Ando,M.;Maeda,H.Arch.Biochem.Biophys. 1992,294,55.
17. Kalyanaraman, B.; Mottley, C.; Mason, R. P. J. Biol.
Chem. 1983, 258, 3855.
18. Chamulitrat, W.; Takahashi, N.; Mason, R. P. J. Biol.
Chem. 1989, 264, 7889.
19. Davies, M. J. Biochim. Biophys. Acta 1988, 964, 28.
20. Davies, M. J. Biochem. J. 1989, 257, 603.
21. Olive, G.; Mercier, A.; LeMoigne, F.; Rockenbauer, A.;
Tordo, P. Free Radical Biol. Med. 2000, 28, 403.
22. Zhang, H.; Joseph, J.; Vasquez-Vivar, J.; Karoui, H.;
Mass spectra were obtained as follows: samples were
diluted in the ratio 1:10.000 in 70% methanol contain-
ing 0.1% formic acid and injected offline to ESI
Q-TOF MS on a Waters Micromass Q-TOF Ultima
Global at a flow rate of 5 ll/min. To acquire appro-
priate spectra for every sample, capillary voltage was
adjusted between 1.2 and 3.0 kV. The mass spectrom-
eter had been previously tuned with [Glu1]-fibrinopep-
tide B to give highest possible sensitivity and a
resolution of 10.000 (FWHM). Mass tuning of the
TOF analyzer was done in the tandem MS mode
using again [Glu1]-fibrinopeptide B. Data analysis
were performed with MassLynx 4.0 SP4 Software
(Waters Micromass).
´
Nsanzumuhire, C.; Martasek, P.; Tordo, P.;
Kalyanaraman, B. FEBS Lett. 2000, 473, 58.
23. Zhao, H.; Joseph, J.; Zhang, H.; Karoui, H.;
Kalyanaraman, B. Free Radical Biol. Med. 2001,
31, 599.
24. O’Brien, P. J. Can. J. Biochem. 1969, 47, 485.