CyDEPMPOs: A class of stable cyclic DEPMPO derivatives with improved properties as mechanistic markers of stereoselective hydroxyl radical adduct formation in biological systems

Article abstract of DOI:10.1016/j.bmc.2011.02.040

The cis/trans diastereoisomeric composition of hydroxyl radical adducts to chiral cyclic nitrones can be used to approach mechanisms of free radical formation in biological systems. Such determination is greatly simplified when both diastereoisomers have ESR spectra with at least two non-overlapping lines. To achieve this prerequisite, a series of DEPMPO-derived spin traps bearing one unsubstituted or alkyl-substituted 2-oxo-1,3,2-dioxaphosphorinane ring were synthesized and their structures were confirmed by X-ray diffraction, ¹H, ¹³C and ³¹P NMR. These CyDEPMPOs nitrones showed variable lipophilicities and LD₅₀ values on murine fibroblasts compatible with a safe use in biological spin trapping. All CyDEPMPOs formed persistent spin adducts with a series of free radicals, including superoxide and hydroxyl (i.e., CyDEPMPOs-OH) and the in vitro half-life times of these two latter were at least as extended as those of parent DEPMPO. Using four methods of CyDEPMPOs-OH formation, the cis-CyDEPMPOs-OH percentage was found significantly varied with substitution on the P-containing ring and, more interestingly, with the generating system.

Full text of DOI:10.1016/j.bmc.2011.02.040

Bioorganic & Medicinal Chemistry 19 (2011) 2218–2230
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
CyDEPMPOs: A class of stable cyclic DEPMPO derivatives with improved
properties as mechanistic markers of stereoselective hydroxyl radical
adduct formation in biological systems
a
a
a,
⇑
b
a
Gaëlle Gosset , Jean-Louis Clément , Marcel Culcasi , Antal Rockenbauer , Sylvia Pietri
a
Sondes Moléculaires en Biologie, Laboratoire Chimie Provence, CNRS UMR 6264, Universités d’Aix-Marseille, Centre Scientifique de Saint-Jérôme (Case 522), 13397
Marseille Cedex 20, France
Chemical Research Center, Institute of Structural Chemistry, Budapest, Hungary
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The cis/trans diastereoisomeric composition of hydroxyl radical adducts to chiral cyclic nitrones can be
used to approach mechanisms of free radical formation in biological systems. Such determination is
greatly simplified when both diastereoisomers have ESR spectra with at least two non-overlapping lines.
To achieve this prerequisite, a series of DEPMPO-derived spin traps bearing one unsubstituted or alkyl-
substituted 2-oxo-1,3,2-dioxaphosphorinane ring were synthesized and their structures were confirmed
Received 21 December 2010
Revised 14 February 2011
Accepted 18 February 2011
Available online 26 February 2011
1
13
31
by X-ray diffraction, H, C and P NMR. These CyDEPMPOs nitrones showed variable lipophilicities and
LD50 values on murine fibroblasts compatible with a safe use in biological spin trapping. All CyDEPMPOs
formed persistent spin adducts with a series of free radicals, including superoxide and hydroxyl (i.e.,
CyDEPMPOs-OH) and the in vitro half-life times of these two latter were at least as extended as those
of parent DEPMPO. Using four methods of CyDEPMPOs-OH formation, the cis-CyDEPMPOs-OH percentage
was found significantly varied with substitution on the P-containing ring and, more interestingly, with
the generating system.
Keywords:
ESR
Spin trapping
DEPMPO
Hydroxyl radical
Diastereoisomeric spin adducts
Ó 2011 Elsevier Ltd. All rights reserved.
ÅÀ
1
. Introduction
Because superoxide (O ) is a key radical species involved in
same experimental conditions for O2 generation, the half-life
2
time was ꢀ15 min for DEPMPO-OOH and only ꢀ1 min for
ÅÀ
2,24
DMPO-OOH.
Despite these advances, in vivo detection of
2
ÅÀ
many biological processes, improving its detection conditions in
aqueous media using electron spin resonance (ESR) spectrometry
has attracted many research groups over the last three decades.
Since the pioneering work of Harbour et al.1 who first reported
O
still remains limited to reductant-free aqueous media (e.g.,
re-suspended cells or sub-cellular cell components
2
2
5
11,26
in phos-
2
phate-based buffers, or isolated organ perfusates ) because of the
poor stability of nitrone-OOH in biological fluids such as blood or
plasma. Recently destructive bioreduction of nitrone-OOH has
been spectacularly slowered, for example, by using b-cyclodextrins
ÅÀ
the ESR spectrum of DMPO-OOH, the protonated O2 spin adduct
on 5,5-dimethyl-1-pyrroline N-oxide (DMPO; Fig. 1A) in water,
synthetic efforts almost uniquely concentrated on the develop-
ment of DMPO-derived, five-membered ring cyclic nitrones that
could form long-lived nitrone-OOH adducts. In this stability regard,
a decisive step was achieved by introducing an electron-withdraw-
2
7,28
either to form inclusion complexes
atized cyclic nitrones.
or as a framework of deriv-
1
3,29,30
However, further use of these sophis-
ticated spin traps in small living animals under oxidant stress,
ideally combined with low-frequency ESR detection, would require
gram scale quantities and toxicity data to be available.
2 2 2
ing group such as –P(O)(OEt) , –CO Et or –CONH at the C-5
position of the DMPO cycle, leading to the 5-diethoxyphospho-
ryl-5-methyl-1-pyrroline N-oxide (DEPMPO), 5-ethoxycarbonyl-
When it does not lead to the total suppression of the ESR signal
(e.g., in the presence of ascorbate where the diamagnetic
N-hydroxylamine is ultimately formed), bioreduction of nitrone-
OOH may result in the formation of nitrone-OH, the hydroxyl
5
-methyl-1-pyrroline N-oxide (EMPO) and AMPO families,
2
–4
respectively (Fig. 1A).
Due to the existence of a strong stabilizing
Å
anomeric effect, increased nitrone-OOH half-lives were reported
radical (HO ) spin adduct. In the DEPMPO and EMPO series, this
in vitro with derivatives of DEPMPO,5–13 EMPO
14–21
or AMPO,
22,23
mechanism has been observed in vitro with glutathione peroxidase
2
,3
which significantly exceeded that of DMPO-OOH. Thus, using the
(GPx), an enzyme found in cytosol, mitochondria and biological
fluids, which converts hydroperoxides into alcohols. Since in
oxygenated medium nitrone-OH can also be formed by metal
⇑
Corresponding author. Tel.: +33 491289 025; fax: +33 491 288 758.
E-mail address: marcel.culcasi@univ-provence.fr (M. Culcasi).
2
,24,31
ion-catalyzed nucleophilic addition of water,
this spin adduct
0
968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.02.040

G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2219
was reported for the C
5
-substituted phenyl DEPMPO5 but this
nitrone showed toxicity upon perfusion in normoxic isolated rat
hearts (unpublished results). More recently, a DEPMPO derivative
in which the P-atom at C-5 is part of a six-membered ring (i.e.,
2
-oxo-1,3,2-dioxaphosphorinane) was reported to yield two ESR
Å
10
distinguishable HO adducts but this behaviour was not further
investigated. In the present paper, we describe the spin trapping
properties of an extended series of such cyclic DEPMPO analogues
(
CyDEPMPOs; Fig. 1B) towards a series of different free radicals.
We particularly focused on the diastereoisomeric content of their
Å
HO adducts obtained using different generators and propose two
of the new nitrones as improved non-toxic spin traps for mecha-
nistic purpose in the study of biological free radicals.
2
2
. Results and discussion
.1. Synthesis
As first introduced in,10 we kept CyDEPMPOs as a general
acronym of cyclic DEPMPO analogues described herein but with
slight modification indicating the nature and number of substitu-
ents in the P-containing ring (Fig. 1B).
Synthetic methods used for the preparation of CyDEPMPOs are
shown in Scheme 1. The 2-oxo-1,3,2-dioxaphosphinanes 2a–2e
were obtained in quantitative yields by modifying the method of
3
3
Page et al. The critical step of the one-pot synthesis involved
addition of 1 equiv of the corresponding 1,3-diol (1a–1e) in THF.
Due to the known anancomeric configurational isomerism at phos-
3
4
phorus in six-membered ring phosphonates, the presence of a
single Me at C-5 in the ring of 2b led to a mixture of cis/trans dia-
stereoisomers with respect to the preferential axial orientation of
the phosphoryl oxygen (Fig. 1C). In this mixture, when homonu-
clear decoupling of each Me at C-5 was performed at either
1
4
54.87 or 582.83 Hz, the H-6 signal in the H NMR spectrum
(
500 MHz in CDCl
diastereoisomer (66%;
while it was a doublet of triplets of triplets for the minor diastereo-
3
) appeared as a triplet of triplets for the major
Figure 1. (A) Structure of cyclic nitrones designed for biological spin trapping; (B)
structure of CyDEPMPOs; (C) cis/trans diastereoisomery in the preferred configu-
ration of the 5-methyl-2-oxo-1,3,2-dioxaphosphorinane framework.
3
3
J5ax,6ax = 11.0 Hz and
J5ax,6eq = 4.3 Hz)
3
3
4
isomer (34%; J5eq,6ax = 6.6 Hz; J5eq,6eq = 3.6 Hz and JP,5eq = 1.5 Hz).
The largest H-5/H-6 coupling appearing in the major diastereoiso-
mer of 2b, indicates an axial–axial orientation for H-5 and H-6 such
as in cis-2b.
is quite often detected in biological spin trapping experiments
involving pyrroline N-oxides and thus cannot be considered as a
Å
reliable marker of a biologically-relevant HO formation, for exam-
The a-aminophosphonates 3a–3e were obtained as solids from
ÅÀ
ple, by a Fenton-like mechanism involving primary O2 generation
the corresponding cyclic phosphites 2a–2e by the previously de-
scribed method involving 2-methyl-1-pyrroline in toluene at
room temperature. The cis/trans-2b mixture, which degraded upon
tentative chromatographic separation, was directly reacted with
followed by its dismutation into hydrogen peroxide (H
2
O
2
). In a re-
35
3
2
cent study, we showed that these three above-mentioned mech-
anisms of nitrone-OH formation can be discriminated through the
ESR analysis of cis/trans diastereoisomers of DEPMPO-OH, being
stereoselectively formed as a result of the presence of a chiral cen-
ter at C-5 of the nitrone (see Fig. 2 with Y = OH). Whereas this al-
2
-methyl-1-pyrroline, yielding cis/trans diastereoisomers of 3b
which could be separated by column chromatography. From the
X-ray crystal structure of 3b (data to be published elsewhere) a
Å
lowed to demonstrate that the HO adduct signal detected in the
effluents of DEPMPO-reperfused isolated rat livers resulted from
ÅÀ
primary O2 formation, cis/trans analysis required accurate simula-
3
2
tion of the two overlapping ESR components of DEPMPO-OH. To
improve the reliability of the cis/trans content in the diastereoiso-
Å
mer method, DEPMPO derivatives having non-overlapping HO ad-
duct spectra (a common feature in EMPO3
,14,16–21
and AMPO
4,22,23
families) would thus be more suitable. Indeed this unique property
Scheme 1. General synthesis of CyDEPMPOs. Reagents and conditions: (i) 1 equiv
H
2
O, 3 h, 70 °C, THF; (ii) 1 equiv diol (1a–1e), 5 h, 70 °C, THF; (iii) 1.1 equiv
2-methyl-1-pyrroline, 4–6 h, rt, toluene; (iv) 2.1 equiv H , 5% Na WO , 48 h, 0 °C,
O.
Figure 2. Diastereoisomers formed upon spin trapping on DEPMPO. Cis/trans
positions of Y ligand are relative to the diethoxyphosphoryl substituent.
2
O
2
2
4
H
2

2
220
G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
Table 1
a
Selected crystallographic data of CyDEPMPOs
1
1
12
O7
3
2
O
1
0
P
4
6
O
5
9
8
N1
O
Compound
Bond lengths (Å)
C(2)–P(6)
Dihedral angles (°)
N(1)–C(5)
P(6)–O(7)
N(1)C(2)P(6)O(7)
C(5)N(1)C(2)C(3)
P(6)O(8)C(9)C(10)
O(12)P(6)O(8)C(9)
4
a
1.301(6)
1.295(4)
1.285(3)
1.284(4)
1.282(4)
1.286(4)
1.839(4)
1.832(3)
1.813(2)
1.836(2)
1.831(3)
1.830(2)
1.464(3)
1.461(2)
1.465(17)
1.459(16)
1.457(2)
1.457(2)
163.9(3)
48.6(2)
À163.86(14)
164.41(15)
À177.52(18)
164.51(18)
À13.2(4)
14.7(3)
À44.1(5)
43.1(3)
46.8(2)
À49.9(2)
28.40(3)
19.5(3)
22.3(3)
À30.9(2)
À23.97(16)
25.47(17)
À7.7(2)
cis-4b
trans-4b
4
4
4
À8.1(2)
c
d
e
15.3(3)
À16.53(3)
12.8(3)
À6.50(2)
a
Values in parentheses are the estimated standard deviations.
trans configuration was found for the major diastereoisomer (65%
of the mixture; Fig. 1C).
may be differences in the spin trapping properties of CyDEPMPOs
in biological milieu (in particular cis/trans ratio) although impor-
tance of solvation in water should be considered.
Nitrones 4a–4e were finally obtained from the corresponding
compounds 3a–3e by a slow addition of H
2 2
O in the presence of
5
% Na WO using water as the solvent instead of methanol as re-
2
4
2.3. Lipophilicity and cytotoxicity of nitrones
3
6
ported. After purification by column chromatography and recrys-
tallization chemical structures of nitrones were fully characterized
by NMR ( H, C and P).
The 1-octanol/water partition coefficient P is an accepted
parameter to model membrane penetration, a higher value indicat-
ing increased lipophilicity of the molecule (a favorable feature for a
spin trap to investigate free radical formation in cell membranes).
With respect to DEPMPO, a significant increase in the P value was
1
13
31
2
.2. X-ray crystal structure analysis of nitrones
Selected crystal parameters for CyDEPMPOs are given in Table 1
and ORTEP drawings with 40% probability displacement thermal
only obtained in the most substituted CyDEPMPOs, i.e., Me
4
Cy-
3
7
DEPMPO and Et CyDEPMPO (Table 2), this latter nitrone being
2
ellipsoids for nitrones cis-4b, 4d and 4e are shown in Figure 3. The
even more lipophilic than 5-diisopropoxyphosphoryl-5-methyl-1-
pyrroline N-oxide (DIPPMPO, see structure in Fig. 1A). In the DEP-
MPO family a dramatic increase of P has been reported (i) with
pyrroline ring adopts an envelope conformation having either a E
(
3
compounds cis-4b, 4c–4e) or a 3E (compounds 4a and trans-4b)
6
form as shown by the signs of the C(5)N(1)C(2)C(3) dihedral an-
gles, leaving the C–P bond in an axial or pseudo-equatorial geom-
etry, respectively. Despite its bulkiness the six-membered
phosphinane ring is oriented over the nitronyl ring in all nitrones,
even the over-crowded 4d and 4e (as shown by N(1)C(2)P(6)O(7)
dihedral angles close to 180°), with the surprising exception of
cis-4b where the free radical trapping site appears much less steri-
cally hindered (Fig. 3). Taken together these data suggest there
increasing the length of the alkyl chain attached to phosphorus,
5
,8
or (ii) in 5-phenylated substituents.
In this study, incubation of 3T3 cells with 50 mM of selected nit-
rones variably affected membrane integrity, as characterized by
cellular LDH release (an indicator of cytotoxicity), and LD50 and
LD10 values were determined from dose–response experiments
(up to 0.1 M; Table 2). Compared to the acyclic N-t-butyl-a-phe-
nylnitrone (PBN), the by far more lipophilic and toxic tested
C4
C4
C4
C5
C3
C2
C3
C3
C5
N1
C5
N1
O
Me
O7
Me
-
-
N1
O-
C2
O
Me
C2
P6
O8
O12
P6
C9
O8
Me
O12
C9
O7
O7
C10
C11
Me
Me
C11
C10
O12
O8
C11
Me
C10
Me
C9
Me
Me
cis-MeCyDEPMPO (cis-4b)
Et CyDEPMPO (4d)
Me CyDEPMPO (4e)
4
2
Figure 3. ORTEP drawings of nitrones cis-4b, 4d and 4e showing the labelling of non-H atoms and their displacement ellipsoids at the 40% probability level. Free radical
addition occurs at C-5 (encircled).

G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2221
Table 2
n-Octanol/water partition coefficients P, molar absorptivity
e and in vitro cell toxicity data of nitrones
Pa
e
b
(L mol cm
À1
À1
)
LDH releasec (IU/L)
LD50 (LD10)d (mM)
Compound
⁄
CyDEPMPO
trans-MeCyDEPMPO
cis-MeCyDEPMPO
0.04 ± 0.03
0.10 ± 0.01
0.10 ± 0.01
0.14 ± 0.01
1.03 ± 0.04
0.23 ± 0.01
0.17 ± 0.01
0.80 ± 0.09
0.19 ± 0.04
15.1 ± 0.76
8050
7.6 ± 1.8
nm
nm
12.5 ± 4.0
25.6 ± 3.7
15.7 ± 5.9
6.4 ± 0.9
122
nm
nm
8306
7849
8260
8020
7992
8168
7500
7807
e
Me
Et CyDEPMPO
Me CyDEPMPO
2
CyDEPMPO
112
⁄
2
55 (31)
93 (68)
129
49 (25)
118
4
e
e
e
⁄
DEPMPO
DIPPMPO
EMPO
⁄
38.9 ± 6.1
11.7 ± 2.7
79.4 ± 13.5
⁄
PBN
17671
7
nm, not measured.
a
Mean value ± SD (n P 4).
Measured at kmax = 234 nm.
b
c
LDH release was measured after 24 h incubation with 50 mM of each spin trap. Values are mean ± SD (n = 6–12). In untreated 3T3 cells baseline LDH release was
⁄
1
2.7 ± 1.9 IU/L. Statistics: p <0.01 versus untreated cells.
d
LD50 and LD10, the concentrations leading to 50% and 10% cell death, respectively, were determined using the neutral red assay from dose–response experiments after
2
4 h incubation with each spin trap. Values are mean ± SD (n = 3).
e
CyDEPMPO, 0.11; DEPMPO, 0.1;6 DIPPMPO, 2.17 and EMPO, 0.1.17
10
Previously reported data were: Me
2
compound, cyclic nitrones were significantly better tolerated by
the cells, the more lipophilic compounds (i.e., Et CyDEPMPO and
(not shown). Bubbling the system with argon prior to photolysis
or after CyDEPMPOs-OOH concentration was allowed to reach a
plateau (ꢀ6–10 min after addition of XO) yielded high resolution
spectra (Fig. 4) which could be accurately simulated assuming a
mixture of cis/trans diastereoisomers with (i) long range hyperfine
coupling constants (hfcs) within the range of that reported for DEP-
2
DIPPMPO) however showing a moderate toxicity. A high spin trap
lipophilicity, by facilitating subcellular distribution,38 could be an
important property in the design of specific anti-free radical pro-
tectors but there have been many controversial examples in cell
3
9,40
41,42
2,32
45
7
studies on PBN
and some EMPO derivatives
where a high
MPO-OOR (R = H,
tBu ) and DIPPMPO-OOH, and (ii) alternate
P value was associated with strong intrinsic cytotoxicity at high
concentrations. With the aim to evidence spin adduct formation
in biological systems under oxidative stress using ESR detection,
LD50 and LD10 data of Table 2 suggest that a safe CyDEPMPOs con-
centration would be at least 30–50 mM. We previously evidenced
the particular involvement of the phosphonate group in the hemo-
dynamic protection against cardiac reperfusion provided by DEP-
linewidth for the major species, due to an intramolecular chemical
exchange between two conformers T1 and T2 in slow chemical
equilibrium (Table 3). From temperature-dependent studies we
have proposed that this latter unique ESR property could reflect a
concerted mechanism involving rotation around the O–O bond
4
6
and pseudo-rotation of the pyrrolidine ring. This phenomenon
leads to a non-uniform broadening of the ESR lines in the trans iso-
43
46
MPO and its pyrrolidine analogue. Thus, a similar property
could explain why DEPMPO and all CyDEPMPOs except Et Cy-
mer causing asymmetry of the signal. Modeling this conforma-
2
tional exchange in CyDEPMPOs-OOR (R = H, tBu) at room
DEPMPO could decrease LDH release of untreated cells. In this re-
gard additional mechanisms are possible: (i) scavenging on the
nitronyl group of radicals produced under usual hyperoxic condi-
temperature yielded mean exchange time values
s within the
nanosecond scale, in the range of that previously reported in the
4
5,46
corresponding DEPMPO peroxyl adducts.
For Me CyDEPMPO-OOH we obtained a more realistic set of
hfcs than that previously reported and, since b-couplings fol-
low dihedral rules, the computed ranges of a and aHb for T1
and T2 rotamers shown in Table 3 are characteristic of a twist
4
1,42
tions of cell culture, as postulated for EMPO-type nitrones,
2
4
4
10
and/or (ii) a known pharmacological effect of a number of pro-
drugs containing the P-containing heterocyle of CyDEPMPOs,
which is also present in cyclic adenosine monophosphate.
P
3
3
45
T
4
or a E envelope for the pyrrolidine ring, respectively. For
2
.4. Superoxide and tert-butoxyl radical adducts
all new nitrones H resolution of the minor cis diastereoisomer,
c
which was not achieved in CyDEPMPOs-OOH (Fig. 4A), was only
observed above 80 °C in cis-DEPMPO-OOH. Conversely all cis-
ÅÀ
46
Although the design of spin traps having O2 adducts with bet-
ter stabilities than available nitrones was not the purpose of this
study, it was nevertheless expected the resulting CyDEPMPOs-
OOH nitroxides would present ESR features and half-life times
close to that of parent DEPMPO-OOH (i.e., ꢀ11–15 min at pH
CyDEPMPOs-OOtBu signals were fully resolved even at room
temperature (Fig. 4B) contrary to cis-DEPMPO-OOtBu.45 For both
types of peroxyl adducts there was no clear change in the stere-
oselectivity of the trapping reaction with the increase of the ste-
ric hindrance of the phosphoryl ligand, the largest yields in cis
2
,6,9
). Generation of O2 or tBuOO Å in the presence of 4a–4e
ÅÀ
ꢀ7
yielded the ESR spectra of CyDEPMPOs-OOH or CyDEPMPOs-OOt-
isomer being unexpectedly found for the crowded Et
(Table 3).
2
CyDEPMPO
Bu, respectively, representative examples being shown in Figure
ÅÀ
4
. Regarding the trapping of O2 the same ESR patterns were ob-
Apparent half-life times (t1/2app) of the superoxide adducts were
determined from a series of repetitive acquisitions using a KO /18-
tained whether the radical was formed using the enzymatic
hypoxanthine (HY)/xanthine oxidase (XO) system in diethylenetri-
aminepentaacetic acid (DTPA)-supplemented aerated phosphate
2
crown-6 ether/DMSO generator in 20 mM phosphate buffer and a
first-order exponential decay approximation. No significant para-
magnetic byproduct was detected during the decay of CyDEPM-
POs-OOH and calibration with DEPMPO gave a t1/2app value of
16.5 min, within the range of previous determinations.2
2
buffer (pH 7.4), or by alternative methods such as KO /crown ether
reagent in DMSO, or photolysis of 30% H . CyDEPMPOs-OOtBu
2 2
O
,6,9,41
adducts were easily obtained by continuous photolysis in pure
tBuOOH. The CyDEPMPOs-OOH signal disappeared when excess
superoxide dismutase (SOD) was added to the HY/XO generator,
while that of CyDEPMPOs-OOtBu was insensitive to the enzyme
4
Among the tested nitrones trans-MeCyDEPMPO and Me CyDEPM-
PO had the more persistent superoxide adducts, about 60% more
compared to DEPMPO (Table 3).

2
222
G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2
Figure 4. High resolution ESR spectra and simulations of (A) superoxide and (B) tert-butoxyl adducts of Et CyDEPMPO. Samples contained: (A) nitrone (0.1 M), hypoxanthine
(
0.4 mM), xanthine oxidase (0.04 units/mL), DTPA (1 mM) incubated for 6.5 min in phosphate buffer (20 mM, pH 7.4) followed by argon bubbling; (B) nitrone (0.05 M) and
tBuOOH (10 mM) continuously UV-photolyzed in degassed anhydrous toluene. Signals were acquired at 9.79 GHz with a 100-kHz modulation frequency and 10 mW
microwave power. Other acquisition parameters (A and B): sweep width (115 G, 100 G); modulation amplitude (0.16 G, 0.05 G); time constant (20.48 ms, 10.24 ms); receiver
5
4
gain (6.3 Â 10 , 6.3 Â 10 ), scan rate (2.7 G/s, 1.2 G/s) and number of accumulated scans (5, 1). Inserts: experimental and calculated spectra of the selected region. Arrows
indicate non-composite lines of the hydroxyl radical adduct (ꢀ9% of the total signal) and (d) indicate selected lines of the cis adduct.
.
2
.5. Spin trapping of H and of C- and O-centered radicals using a
the lines were symmetric indicating the presence of one species
and calculated a values of 45–54 G suggest a preferential pseu-
Å
primary HO source or photolysis (Table 4)
P
do-axial geometry for the C–P bond with a kinetically-favoured
trans configuration. Stereoselective trapping of these four radicals
on nitrones from the DEPMPO family has been previously re-
4
Reduction of CyDEPMPOs by NaBH in the presence of molecu-
lar oxygen afforded the corresponding CyDEPMPO-H nitroxides.
After careful degassing of the samples, highly resolved signals were
obtained showing two non-equivalent b-hydrogens and hfcs simi-
2
,6,7
ported,
with the exception of C
5
-substituted phenyl DEPMPO
ÅÀ
which yielded cis/trans mixtures of diastereoisomers with CO
and tBuO having low a
torial geometry of the diethoxyphosphoryl group.
When the glutathionyl radical (GS ) was generated in the pres-
ence of CyDEPMPOs all resulting ESR spectra were composed of
mixtures of two nitroxides, the simulation of which gave sets of
2
Å
lar to that obtained in DEPMPO-H (i.e., in G): a
Hb1 = 21.24, aHb2 = 20.31 and a = 0.33 (Me); 0.53, 0.45, 0.41,
.19).
Other spin adducts of C-centered radicals were obtained in
N
= 15.31, a
P
= 50.52,
P
values of ꢀ35 G, indicating a pseudo-equa-
5
a
0
Hc
.
phosphate buffer either using excess adequate substrate in the
Å
47
presence of Fenton-generated HO as a primary source, or photolyt-
hfcs close to that of DEPMPO-SG for the major species, and Me2-
ÅÀ
Å
10
ically. For methyl, 1-hydroxyethyl, CO2 and tBuO radical adducts,
P
CyDEPMPO-OH (e.g., with a >50 G; see also data below) for the

G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2223
Table 3
Computer simulated ESR parameters (upper part) and apparent half-life times (lower part) of superoxide and tert-butylperoxyl adducts of CyDEPMPOs at room temperature
Radical
hfcs
G)
CyDEPMPO
(90%)c
trans-MeCyDEPMPOa cis-MeCyDEPMPOa
Me
2
CyDEPMPO
Et
2
CyDEPMPO
4
Me CyDEPMPO
(
b
Å
trans -HOO
(87%)
(96%)
(95%)
(88%)
(93%)
T2(55%) T1(56%)
d
T1(45%)
T2(55%) T1(40%)
T2(60%) T1(50%)
T2(50%) T1(47%)
T2(53%) T1(45%)
T2(44%)
e
s
= 6.1
s
= 5.5
s
= 6.5
s
= 19.6
s
= 6.5
s = 5.4
a
a
a
a
N
13.2
57.0
12.8
0.35 (Me)
13.0
46.4
9.0
12.6
56.7
13.1
0.35 (Me)
0.90/0.63
0.45/0.33
13.3
47.1
9.2
12.9
56.9
12.9
0.37 (Me)
0.91/0.62
0.44/0.44
13.2
47.0
9.0
12.8
53.9
12.1
0.33 (Me)
0.85/0.64
0.47/0.11
13.2
49.5
9.7
13.2
56.9
13.0
0.38 (Me)
0.87/0.64
0.46/0.34
12.8
47.3
9.1
12.8
56.6
12.3
0.39 (Me)
0.6/0.8
0.2
13.3
46.6
9.0
P
Hb
H
c
0
0
.91/0.63
.47/0.45
cis-HOOÅ
(10%)
13.4
40.6
11.2
1.41
(13%)
13.4
40.8
9.4
(4%)
13.5
39.3
11.6
1.83
(5%)
13.4
40.7
9.1
(12%)
13.4
40.7
9.2
(7%)
13.6
43.1
10.6
1.30
a
a
a
a
N
P
Hb
H
c
1.70
1.74
1.57
trans-tBuOOÅ
(92%)f
(92%)
(91%)
(86%)
(88%)
(92%)
T1(58%)d
T2(42%) T1(58%)
= 38
T2(42%) T1(65%)
= 60
T2(35%) T1(57%)
= 12
T2(43%) T1(57%)
= 12
T2(43%) T1(56%)
= 25
T2(44%)
s
= 18
s
s
s
s
s
a
a
a
a
N
P
12.35
48.09
9.28
0.38 (Me)
12.34
52.27
10.53
12.36
49.75
10.37
0.36 (Me)
1.12/0.58
0.47/0.43
12.42
47.84
8.57
12.33
47.59
9.11
0.36 (Me)
1.08/0.60
0.47/0.43
12.05
53.09
10.55
12.39
47.64
9.13
0.36 (Me)
1.08/0.60
0.47/0.43
12.14
53.20
10.57
12.35
47.66
9.14
0.36 (Me)
1.08/0.60
0.47/0.43
12.05
53.15
10.55
12.26
52.83
10.16
0.38 (Me)
1.04/0.62
0.49/0.48
12.20
48.15
9.23
Hb
H
c
1
0
.10/0.59
.46/0.46
cis-tBuOOÅ
(8%)
(8%)
(9%)
(14%)
(12%)
(8%)
a
a
a
a
N
P
12.75
39.20
8.29
12.77
37.46
7.84
0.37 (Me)/
1.84
12.70
40.70
8.40
0.45 (Me)/
1.70
12.82
37.77
7.85
0.37 (Me)/
1.85
12.76
37.68
7.70
0.37 (Me)/
1.88
12.81
40.09
9.38
Hb
H
c
0.38 (Me)/
1.58
1
0
.71
.54/0.45
0.51/0.46/
0.50/0.30
0.51/0.46/
0.16
0.51/0.46/
0.16
0.15
t
1/2app of
10.2
26.3
16.2
12.3
12.4
26.4
superoxide
adductg
min)
(
a
b
c
d
e
f
Relative position of P = O versus Me substituent of the P-containing heterocycle.
cis/trans indicate the relative position of trapped radical with respect to the P-containing heterocyle.
From HY/XO system in phosphate buffer (pH 7.4).
T1 and T2 refer to slowly exchanging rotamers.
Computed average exchange time (ns).
From t-BuOOH/h
m
in degassed toluene.
g
From KO
2
/crown ether/DMSO generator in 20 mM phosphate buffer (pH 7.4).
minor species. A very good agreement with experimental signals
r >0.90) was obtained due to the distinct separation of the ESR
calculated from the best resolved signals recorded in deoxygenated
milieu.
(
low-field lines from each adduct, ranging 2.0–3.0 G (Fig. 5A) and
we assume they are cis/trans diastereoisomers. Likewise, when
the spectra were recorded using the same conditions the relative
amount of the minor species decreased with increasing substitu-
tion on the dioxaphosphorinane ring, a property compatible with
the formation of the cis diastereoisomer.
For each pair of diastereoisomers, couples of calculated aHb (i.e.,
ꢀ14 G, 12 G) matched those we previously calculated for DEPMPO-
OH and DIPPMPO-OH based on the analysis of their ESR line asym-
metries and, due to this similarity, we therefore assigned them to
the cis and trans configurations of CyDEPMPOs-OH, respectively
(Table 4). Accordingly, b-couplings values in trans-CyDEPMPOs-
OH indicate pseudo-axial positions for both hydroxyl and phos-
phoryl groups, with dihedral angles lower than 18° and 30°,
respectively, a geometry which should favor stabilizing anomeric
interactions. Such an electronic effect may also explain the rela-
3
2
2
.6. Diastereoisomery and kinetic behavior of hydroxyl radical
adducts
Å
For evaluation of HO trapping ability and the stability of the
resulting CyDEPMPOs-OH adducts we first used either a H
P
tively high a values (by about 2 G compared to trans adducts) in
most cis-CyDEPMPOs-OH, which nevertheless retain their b-C–H
bond in a less pseudo-equatorial position, giving higher aHb cou-
2
O
2
2
+
(
1
2 mM)/Fe (1 mM)-driven Fenton reaction or UV photolysis of
% H as radical producing systems. As expected from earlier re-
sults on Me
CyDEPMPO-OH,10 these generators yielded mixtures
of cis/trans diastereoisomers having very distinct ESR signals
Fig. 5B). Even in non degassed milieu the tops of the low-field
lines from the spectra of each diastereoisomer were separated by
.5–1.9 G. This unique spectral property of CyDEPMPOs-OH al-
2
O
2
P 4
plings. From this rationale the marked lower a value for cis-M Cy-
2
DEPMPO-OH could reflect an increased steric repulsion between –
OH and –P(O) groups that force the latter to occupy a more equa-
torial position (Table 4).
To evaluate the usefulness of new CyDEPMPOs as tools to inves-
tigate the mechanism of hydroxyl radical adduct formation in a
biological environment, we determined cis/trans ratios of CyDEPM-
POs-OH using either the two above-cited generators, or two
(
1
lowed to perform more reliable simulations because less parame-
ters had to be adjusted, and hfcs listed in Table 4 were then

2
224
G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
Figure 5. (A) ESR spectrum of glutathionyl radical adduct of CyDEPMPO obtained upon incubation of 0.1 M nitrone, 0.13 M GSH and a Fenton reagent consisting of 1 mM
FeSO and 1 mM H in 20 mM phosphate buffer. Instrument settings: microwave power, 10 mW; sweep width, 120 G; modulation amplitude, 0.07 G; time constant,
0.48 ms, receiver gain, 5 Â 10 and scan rate 1.4 G/s. The low-field lines of trans (s) and cis (d; 13% of the total signal from the simulated signal shown below)
diastereoisomers are shown. (B) Typical ESR spectra in 20 mM phosphate buffer (pH 7.4) of hydroxyl radical adducts of (a) CyDEPMPO (from photolysis of 1% H ); (c)
CyDEPMPO (from Fe3 (1 mM)-catalyzed nucleophilic addition of water); (g)
CyDEPMPO (from reduction of superoxide adduct by GSH+GPx). From simulations (b, d, f and h) of signals (a, c, e and g), respectively, the calculated percentages of cis (d)
4
2 2
O
4
2
2 2
O
+
Et
Et
2
4 2 2 4
CyDEPMPO (from a Fenton reaction between 1 mM FeSO and 1 mM H O ); (e) Me
2
over trans (s) diastereoisomers are: 33, 51, 41, and 26, respectively. Microwave power was 10 mW throughout and other settings were (signals a/c/e/g): time constant (ms),
5
4
0.96/40.96/20.48/20.28; modulation amplitude (G), 0.05/0.12/0.05/0.5; receiver gain (Â10 ), 3.2/5/1.6/1; scan rate (G/s), 1.2/0.7/1.4/2.
additional radical- and nonradical-based systems, that is, reduction
of pre-formed superoxide adduct by a mixture of GPx and reduced
glutathione (GSH), or nucleophilic addition of water, respectively.
In order to reliably compare all nitrones experiments were per-
formed by carefully controlling time since we have shown that dia-
stereoisomers of DEPMPO-OH and DIPPMPO-OH have different
spin adducts were allowed to build up for 1 min then addition of
catalase (20 units/mL) or cut off of UV irradiation were performed,
respectively. In these two systems spectra were recorded 1.5 min
after HO generation. In experiments using Fe (1 mM)-catalyzed
nucleophilic addition of water, the maximal concentration of spin
adduct was considered to occur immediately and ESR scan was ini-
tiated after 1 min.
Å
3+
3
2
stabilities.
Briefly, in the reductive system the HY/XO superoxide generat-
ing system was run for 6 min in nitrone (0.1 M)- and DTPA (1 mM)-
The mean percentages of cis-CyDEPMPOs-OH along with the
four generating systems are given in Table 5. Of the studied mech-
anisms leading to formation of the HO spin adduct, nucleophilic
Å
containing phosphate buffer (20 mM; pH 7.4), then formation of
ÅÀ
O
and its potential dismutation product (H
2
O
2
) was immediately
addition of water under metal ion catalysis showed the lowest ste-
reoselectivity with little differences between nitrones. The mecha-
nism involves the concerted formation of a metal ion complex with
the negatively charged nitronyl oxygen and simultaneous nucleo-
philic attack of one oxygen lone pair of water on the resulting elec-
2
stopped by adding 10 units/mL SOD and 20 units/mL catalase. Con-
version of CyDEPMPOs-OOH into CyDEPMPOs-OH was then
achieved by adding GSH (1.2 mM) and GPx (10 units/mL) to the
mixture so that all spectral acquisitions were started 7 min after
initial addition of XO. In the Fenton and photolytic systems, the
3
1
tron-poor nitronyl carbon. Although it is not a radical pathway,

G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2225
Table 4
Calculated ESR parameters of different radical adducts of CyDEPMPOs
Radical
HÅa
hfcs (G) CyDEPMPO
trans-MeCyDEPMPO
cis-MeCyDEPMPO
Me
2
CyDEPMPO
Et
2
CyDEPMPO
4
Me CyDEPMPO
a
a
a
a
N
15.23
52.40
20.89/20.53
0.42 (Me)
15.21
52.08
20.91/20.46
0.41 (Me)
0.50/0.35
0.32/0.18
15.17
52.67
20.97/20.34
0.40 (Me)
0.49/0.39
0.34/0.18
15.18
52.70
21.00/20.36
0.42 (Me)
0.52/0.38
0.30/0.19
15.16
52.94
21.06/20.26
0.43 (Me)
0.51/0.35
0.32/0.18
15.28
53.34
21.17/20.17
0.40 (Me)
0.51/0.51
0.32/0.18
P
Hb
H
c
0
0
.50/0.34
.33/0.18
b
ÅCH
3
a
a
a
N
15.04
49.44
21.34
15.03
49.01
21.33
15.02
49.89
21.29
14.23
48.15
19.91
14.98
50.00
20.91
14.25
49.22
19.38
P
Hb
c
ÅCH(OH)CH
3
a
a
a
N
14.54
50.82
20.87
14.53
50.26
20.87
14.52
51.40
20.83
14.12
50.72
20.13
14.44
51.44
20.63
14.13
52.10
19.84
P
Hb
Å
d
CO À
a
a
a
N
14.37
53.46
16.76
14.37
53.05
16.80
14.39
53.77
16.82
14.32
53.47
16.73
14.32
53.94
16.60
14.38
54.01
16.72
2
P
Hb
GSÅe
trans
cis
trans
cis
trans
cis
trans
cis
trans
cis
trans
cis
(
87%)
(13%) (88%)
14.06 14.06
51.57 46.97
16.15 14.49
(12%) (87%)
14.05 14.07
50.80 47.73
15.88 14.41
(13%)
(92%)
(8%)
(96%)
(4%)
(93%)
(7%)
a
a
a
N
14.06
47.30
14.43
14.09 14.03
52.21 47.56
16.26 14.36
14.04 14.01
51.29 47.73
16.00 14.34
13.86 14.06
53.47 48.10
14.25 13.95
14.15
52.74
15.88
P
Hb
tBuOÅf
a
a
a
a
12.87
45.23
7.99
0.43 (Me)
12.95
44.59
7.90
12.75
46.23
7.88
12.85
45.53
7.97
12.86
45.68
8.07
12.70
46.70
8.04
N
P
Hb
H
c
0.43 (Me)
0.45 (Me)
0.43 (Me)
0.43 (Me)
0.44 (Me)
1
0
.31/0.81
.28/0.22
1.34/0.81
0.29/0.22
1.29/0.83
0.32/0.22
1.29/0.82
0.28/0.23
1.27/0.81
0.32/0.23
1.22/0.81
0.26/0.19
trans-HOÅg
a
a
a
a
13.96
48.78
12.42
0.37 (Me)
13.91
49.06
12.26
0.36 (Me)
0.66/0.36
0.30/0.11
13.94
49.22
12.45
0.33 (Me)
0.59/0.53
0.18/0.03
13.93
48.90
12.30
0.42 (Me)
0.58/0.26
0.20/0.12
13.93
49.00
12.36
0.38 (Me)
0.68/0.39
0.23/0.22
13.98
49.73
11.97
0.99
N
P
Hb
H
c
0
0
.60/0.30
.22/0.15
cis-HOÅg
a
a
a
a
13.98
50.35
14.31
0.37 (Me)
13.90
50.86
14.15
0.40 (Me)
0.50/0.37
0.29/0.12
14.03
51.09
14.18
0.40 (Me)
0.53/0.47
0.18/0.14
13.93
50.53
14.11
0.42 (Me)
0.72/0.48
0.25/019
13.98
50.93
13.84
0.46 (Me)
0.59/0.34
0.27/0.08
13.98
48.80
14.15
N
P
Hb
H
c
0
0
.72/0.55
.23/0.18
a
b
c
d
e
f
Fenton reagent supplemented with aNaBH
4
in water.
2
2
0
0
0% DMSO.
0% EtOH.
.05 M ammonium formate.
.13 M GSH.
Photolysis of 10 mM (tBuO)
2
in degassed toluene.
g
Fenton reaction in degassed 20 mM phosphate buffer.
Table 5
a
Spin trapping of hydroxyl radical on CyDEPMPOs: comparison of the percentage of the cis diastereoisomer as a function of the producing system and apparent half-life times of
spin adductsb
Nitrone
Photolysis of H
2
O
2
Fenton reaction
Enzymatic reduction
of superoxide adduct
Nucleophilic addition
t
1/2app (min)
CyDEPMPO
trans-MeCyDEPMPO
cis-MeCyDEPMPO
31
41
29
35
16
19
37
48
41
42
44
7
21
24
26
25
22
11
42
52
44
43
43
41
46
3.9
5.7
5.4
3.8
4.2
Me
Et CyDEPMPO
Me CyDEPMPO
DEPMPO
2
CyDEPMPO
2
4
11.6
9.0
c
c
c
34
35
25
a
Extracted from computer simulations of ESR spectra recorded under time controlled conditions as described in Section 2. The data represent means of 2–3 determinations
(
in 20 mM phosphate buffer; pH 7.4) and are accurate within 5%.
b
From decay curves recorded using the catalase-arrested, DTPA-free Fenton system as the radical generator.
c
32
Data from Culcasi et al. obtained under the same experimental conditions.
nucleophilic addition can play an important role in the study of
biological spin trapping using nitrones because trace metal ions
can be released from damaged cells in inflammatory conditions
centage for nucleophilic addition in vitro is very different from that
of the other tested mechanisms (Fig. 6 and Table 5), the nitrone
may be probe of biological mechanisms involving the release of
metal ions.
such as rheumatoid arthritis.48 Since cis-Me
CyDEPMPO-OH per-
4

2
226
G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
systems under oxidative stress led us to synthesize a series of
new DEPMPO derivatives containing a 2-oxo-1,3,2-dioxaphosphor-
inane ring. All prepared CyDEPMPOs nitrones were isolated and
characterized as stable crystals, and their low toxicities on cultured
murine fibroblasts should be compatible with biological spin trap-
ping. We have recently proposed that valuable mechanistic infor-
mation would be gained from the determination of relative cis/
trans content of hydroxyl radical spin adducts provided each diast-
3
2
reoisomer will have distinct ESR lines. All tested CyDEPMPOs
meet this claim in vitro, that is, the diastereoisomeric ratios were
significantly different using biologically-relevant pathways of
hydroxyl radical adduct formation. Importantly, synthesized
CyDEPMPOs were of different lipophilicities compared to DEPMPO
and also yielded relatively persistent superoxide spin adducts. We
believe CyDEPMPOs represent a fruitful class of spin traps to inves-
tigate free radical generation in vivo.
Figure 6. ESR spectra of the Me
Fenton reaction (from 0.1 M nitrone, 1 mM H
phosphate buffer) or (b) nucleophilic synthesis (from 0.1 M nitrone and 1 mM FeCl
in water). Signals were recorded 1 min after (a) addition of excess catalase
following a 3 min-incubation of the reactants, or (b) addition of iron. Computer
simulations indicate that cis-Me CyDEPMPO-OH (see structure) accounts for (a) 7%
4
and (b) 41% of the total signal with respect to the trans diastereoisomer,
respectively.
4
CyDEPMPO hydroxyl radical adduct formed by (a) a
2
O
2
and 2 mM FeSO in 20 mM
4
3
4
4
. Experimental
When CyDEPMPOs-OH is the result of radical addition to the
nitrone, the cis/trans ratio is a function of steric and electronic
factors that drive the trapping reaction and of the relative stability
of both diastereoisomers. In the DEPMPO family stereoselectivity
of the spin trapping reaction is expected due to the steric hin-
.1. Reagents and analytical instrumentation
Nuclear magnetic resonance (1H NMR at 300.1 MHz, 13C NMR at
31
7
5.5 MHz and P NMR at 121.5 MHz) spectra were recorded using
drance of the phosphorylated versus the methyl substituent, which
a Bruker AVL 300 spectrometer. Chemical shifts are expressed in d
ppm relative to internal tetramethyl silane ( H and C) or external
8
abbreviations s, d, t, m and q refer to singlet, doublet, triplet, mul-
tiplet and quartet signal, respectively. Melting points were deter-
mined using a B-540 Buchi apparatus and are uncorrected.
Elemental analyses were performed using a Thermo Finnigan
EA-1112 analyzer and were within 0.2% of the theoretical values.
Column chromatography was performed on Merck Silica Gel 60
should favor the trans diastereoisomer (Fig. 2).5
,7,9
In agreement
1
13
with their ‘folded’ crystal structure (Fig. 3), the two more bulky nit-
rones (namely Et CyDEPMPO and Me CyDEPMPO) yielded the low-
31
3 4
5% H PO ( P) and coupling constants J are given in Hertz. The
2
4
.
est cis percentages when HO was produced photolytically. Despite
its X-ray structure, in protic solvents, due to solvation, the two
faces of cis-MeCyDEPMPO should not be equally accessible. On
the other hand, the lower stereoselectivity observed for the trap-
ping of hydroxyl radical is in agreement with the lower steric hin-
drance of the trapping sites (Table 5).
(
230–400 mesh). Spectrophotometrical measurements were per-
In superoxide spin trapping experiments in aqueous milieu, the
metal ion chelator DTPA is often introduced to prevent traces met-
als present in commercial phosphate buffers from catalyzing hy-
2
formed on a Saphas UVmc spectrophotometer (Monaco). Solvents
and reagents were of the highest grade available from commercial
suppliers and were used without further purification except for 2-
methyl-1-pyrroline (from Acros Organics, Halluin, France) which
was distilled under reduced pressure (20 mm Hg, bp: 50 °C) prior
to use. Commercial diols used in the syntheses of phosphites
2
4
droxyl radical adduct formation through nucleophilic addition.
In our tested Fenton reagent, DTPA was omitted to be more rele-
vant to biological conditions and because it increases the ESR line-
3
2
width.
Under these conditions, the data of Table 5 are
2
2
a–2e were 1,3-propanediol (1a), 2-methyl-1,3-propanediol (1b),
,2-dimethyl-1,3-propanediol (1c), 2,2-diethylpropane-1,3-diol
characteristic of a diastereoselective trapping reaction, in line with
32
previous results on DEPMPO. According to Table 5, addition of
(1d) and 2,4-dimethylpentane-2,4-diol (1e). The nitrones DEP-
Å
HO on Me
4
CyDEPMPO is highly stereoselective. In vivo metal cat-
35
7
3
MPO, DIPPMPO and EMPO were synthesized and purified as
previously described. PBN was purchased from Sigma–Aldrich
Å
alysts that are potential triggers of HO formation from H
2
O
2
are
susceptible to chelate many biological targets including proteins,
(
Saint-Quentin Fallavier, France). Medium and reagents used in cell
Å
favoring a site-specific rather than a radiomimetic HO genera-
culture were from Life Technologies Inc. (Gaithersburg, MD, USA).
4
9
tion. Thus, differences in cis/trans ratios between the photolytical
Å
and the Fenton HO generators in this study may be due to different
4.2. Synthesis
iron complexing properties of the nitrones.
The cis percentages of CyDEPMPOs-OH in the reductive system
4.2.1. Synthesis of phosphites 2a–2e
were within the range of that of DEPMPO,32 except for Me
4
Cy-
Water (540 mg, 30.6 mmol) was added to HMPT (5 g,
30.6 mmol) in 50 mL refluxing anhydrous THF under N2 atmo-
sphere. The mixture was stirred at reflux for 2 h and 30.6 mmol
of the corresponding diol (1a–1e) in 20 mL of dry THF was added
dropwise. After 2 h reflux, the mixture was concentrated under re-
duced pressure to give the corresponding cyclic phosphite (2a–2e),
DEPMPO which again yielded the lowest value (Table 5). This is an-
other advantage of this nitrone in our search for a sensitive tool to
investigate bio-radical mechanisms by ESR.
The apparent half-lives of CyDEPMPOs-OH have been deter-
mined by using the DTPA-free, catalase-arrested Fenton system
from 0.1 M buffered solutions of nitrones. In these conditions most
hydroxyl radical adducts gave t1/2app values below 6 min, being sig-
3
1
the purity of which was found to be >95% by P NMR, with d val-
50
ues (in CDCl ) in agreement with literature data: 3.47 (2a), 3.21
3
5
1
52
31
nificantly shorter than that of DEPMPO-OH and Me
4
CyDEPMPO-
(2c) and 4.28 (2d). Other non-reported P NMR d values in
CDCl3 were: 3.97 (2e) and 2.16 and 2.98 for trans and cis isomers
of 5-methyl-2-oxo-1,3,2-dioxaphosphorinane (2b), respectively
OH (Table 5). Using the reductive system we reported an apparent
32
half-life of ꢀ7.5 min for DEPMPO-OH.
. Conclusions
(
see assignment in Section 2).
3
4
.2.2. Synthesis of aminophosphonates 3a–3e
Our interest in the development of spin traps that can give in-
sights on the formation of hydroxyl radical adducts in biological
A mixture of 2-methyl-1-pyrroline (2.8 g, 33.7 mmol) and the
corresponding cyclic phosphite 2a–2e (30.6 mmol) were stirred

G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2227
in toluene (5 mL) for 4–6 h at room temperature (conversion was
CH
2 3
CH ), 1.66–1.92 (m, 4H, 3-H, 4-H, NH), 2.21–2.36 (m, 1H, 3-
3
1
checked by TLC or P NMR), the mixture was poured into 30 mL
water, slowly acidified to pH 3 with 11 N HCl then quickly ex-
tracted with tert-butyl methyl ether (TBME) (3 Â 20 mL). The aque-
ous phase was basified to pH 9–10 with NaOH pellets then
H), 2.85–2.97 (m, 1H, 5-H), 3.05–3.13 (m, 1H, 5-H), 4.10–4.23 (m,
1
3
4H, OCH
22.52 (CH
(d, J = 4.5 Hz, 4-C), 34.80 (d, J = 3.8 Hz, 3-C), 37.55 (d, J = 5.3 Hz,
CEt ), 47.36 (d, J = 6.8 Hz, 5-C), 60.44 (d, J = 163.1 Hz, 2-C), 73.69
(d, J = 7.5 Hz, OCH ), 73.79 (d, J = 7.5 Hz, OCH ); Anal. Calcd for
2
); C NMR (CDCl
3 2 3 2 3
) d 7.02 (CH CH ), 7.11 (CH CH ),
2
CH ), 22.99 (CH CH
3
2
3
), 24.32 (d, J = 7.5 Hz, 2-Me), 25.88
extracted with CH
were dried over MgSO
crude aminophosphonates 3a–3e as white powders which were
then purified by SiO column chromatography with eluent
CH Cl /EtOH 8/1.
2
Cl
2
(4 Â 20 mL). The combined organic phases
2
4
, filtered and roto-evaporated to give the
2
2
C
12
H24NO
3
P: C, 55.17; H, 9.20; N, 5.36. Found: C, 55.25; H, 9.56;
2
N, 5.28.
2
2
4
.2.2.6.
2-Methyl-2-(4,4,6,6-tetramethyl-2-oxo-1,3,2-diox-
4
.2.2.1. 2-Methyl-2-(2-oxo-1,3,2-dioxaphosphinan-2-yl)pyrroli-
aphosphinan-2-yl)pyrrolidine (3e). White crystals. Yield: 63%.
31
31
1
dine (3a). White crystals. Yield: 52%. Mp 79 °C. P NMR (CDCl
3
) d
) d 1.39 (d, J = 15 Hz, 3H, 2-Me), 1.61–2.10
, 3-H, 4-H, NH), 2.21–2.33 (m, 1H, 3-H),
.85–2.93 (m, 1H, 5-H), 3.05–3.11 (m, 1H, 5-H), 4.42–4.54 (m,
H, OCH –CH –CH ) d 24.24 (d, J = 6.8 Hz, 2-
O); 13C NMR (CDCl
–CH –CH ),
4.77 (d, J = 3.8 Hz, 3-C), 47.35 (d, J = 3.8 Hz, 5-C), 60.38 (d,
J = 163.8 Hz, 2-C), 67.30 (d, J = 8.3 Hz, CH –CH –CH ), 67.40 (d,
J = 8.3 Hz, CH –CH –CH ); Anal. Calcd for C P: C, 46.83; H,
.86; N, 6.83. Found: C, 46.33; H, 7.83; N, 6.86.
Mp 93 °C. P NMR (CDCl
J = 18 Hz, 3H, 2-Me), 1.44 (3H, CCH
CCH ), 1.70–1.90 (m, 4H, 3-H, 4-H, NH), 1.97–2.13 (dd, J = 15 Hz,
J = 6 Hz, 2H, C(CH CH C(CH ), 2.21–2.38 (m, 1H, 3-H), 2.96–
3.03 (m, 2H, 5-H); C NMR (CDCl ) d 24.30 (d, J = 7.5 Hz, 2-Me),
25.64 (d, J = 5.3 Hz, 4-C), 30.89–31.19–31.30–31.37 (C(CH
CH C(CH ), 34.63 (d, J = 3.8 Hz, 3-C), 47.05 (d, J = 6.8 Hz, 2H, 5-C),
47.32 (d, J = 8.3 Hz, C(CH CH C(CH ), 59.33 (d, J = 163.1 Hz, 2-
C), 79.73 (d, J = 6.8 Hz, OCMe ), 79.82 (d, J = 6.8 Hz, OCMe ); Anal.
Calcd for C12 P: C, 55.16; H, 9.26; N, 5.36. Found: C, 54.77;
H, 9.28; N, 5.31.
3
) d 25.32; H NMR (CDCl
3
) d 1.33 (d,
1
2
6.00; H NMR (CDCl
3
3
), 1.45 (3H, CCH
3
), 1.59 (6H,
(
m, 6H, CH –CH –CH
2
2
2
3
2
4
3
)
2
2
3 2
)
13
2
2
2
3
3
Me), 25.90 (d, J = 3.8 Hz, 4-C), 26.76 (d, J = 8.3 Hz, CH
2
2
2
3 2
) -
3
2
3
2
2
2
3
)
2
2
3 2
)
2
2
2
8
H20NO
3
2
2
7
H24NO
3
4
.2.2.2. trans-2-Methyl-2-(5-methyl-2-oxo-1,3,2-dioxaphosph-
inan-2-yl)pyrrolidine (trans-3b). White crystals. Yield: 55%. Mp
4.2.3. Synthesis of nitrones 4a–4e
To aminophosphonates 3a–3e (24.5 mmol) in water (50 mL)
was added at 0 °C during 48 h under stirring, 30%
(51.45 mmol, 4.4 mL) in presence of a catalytic amount of Na WO
2
(403 mg, 1.22 mmol). At the end of the addition, aqueous phase
3
1
1
9
3
1 °C. P NMR (CDCl
H, CHCH ), 1.40 (d, J = 15 Hz, 3H, 2-Me), 1.62–1.90 (m, 4H, 3-H, 4-
), 2.79–
.87 (m, 1H, 5-H), 3.08–3.15 (m, 1H, 5-H), 4.16–4.38 (m, 4H,
3
) d 23.25; H NMR (CDCl
3
) d 0.88 (d, J = 6 Hz,
3
H
2
O
2
H, NH), 2.22–2.32 (m, 1H, 3-H), 2.40–2.50 (m, 1H, CHCH
2
OCH
3
4
1
3
2
CHCH
3
CH
2
O); C NMR (CDCl
3
) d 11.95 (CHCH
3
), 24.31 (d,
was saturated with NaCl, extracted first with TBME (2 Â 50 mL)
J = 6.8 Hz, 2-Me), 25.98 (d, J = 3,8 Hz, 4-C), 30.96 (d, J = 9 Hz,
CHCH ), 34.97 (d, J = 2.8 Hz, 3-C), 47.45 (d, J = 6.8 Hz, 5-C), 61.31
d, J = 166.8 Hz, 2-C), 73.82 (d, J = 8.3 Hz, OCH ), 73.96 (d,
J = 7.5 Hz, OCH ); Anal. Calcd for C P: C, 49.31; H, 8.28; N,
.39. Found: C, 49.33; H, 8.52; N, 6.36.
then with CHCl
3
(4 Â 50 mL). The combined chloroform phases
3
were dried over MgSO
to lead to crude nitrones 4a–4e that were first purified by SiO
umn chromatography with eluent CH Cl /EtOH 7/1 then recrystal-
lized in TBME/CH Cl These separation and purification
4
, filtered and concentrated under vacuum
(
2
2
col-
2
9
H18NO
3
2
2
6
2
2
.
procedures allowed to separate cis/trans diastereoisomers of 4b,
the stereochemistry of which was determined by X-ray diffraction
(see Section 2).
4
.2.2.3.
cis-2-Methyl-2-(5-methyl-2-oxo-1,3,2-dioxaphosph-
3
1
inan-2-yl)pyrrolidine (cis-3b). Yellow oil. Yield: 16%. P NMR
1
(
CDCl
3
) d 27.7; H NMR (CDCl
.40 (d, J = 18 Hz, 3H, 2-Me), 1.60–1.93 (m, 4H, 3-H, 4-H, NH),
.22–2.38 (m, 2H, 3-H, CHMe), 2.94–3.01 (m, 1H, 5-H), 3.03–3.12
CHMeCH
O); 13C NMR (CDCl
), 23.90 (d, J = 7.5 Hz, 2-Me), 25.56 (d, J = 5.3 Hz, 4-
), 34.57 (d, J = 3.8 Hz, 3-C), 47.10 (d,
J = 6.8 Hz, 2H, 5-C), 59.99 (d, J = 163 Hz, 2-C), 70.80 (d, J = 6.8 Hz,
OC(CH ), 70.99 (d, J = 6.8 Hz, OC(CH ); Anal. Calcd for
3 3
) d 0.90 (d, J = 6 Hz, 3H, CHCH ),
1
2
4.2.3.1.
2-Methyl-2-(2-oxo-1,3,2-dioxaphosphinan-2-yl)-3,4-
dihydro-2H-pyrrole 1-oxide [CyDEPMPO (4a)]. Yellow crystals.
31
1
(
m, 1H, 5-H), 4.12–4.28 (m, 4H, OCH
2
2
3
)
Yield: 90%. Mp 115 °C; P NMR (CDCl
d 1.73 (d, J = 15 Hz, 3H, 2-Me), 1.82–1.89 (m, 2H, CH
2.24–2.35 (m, 1H, 3-H), 2.56–2.80 (m, 2H, 4-H), 2.90–3.03 (m,
1H, 3-H), 4.33–4.54 (m, 3H, O–CH –CH –CH –O), 4.98–5.07 (m,
–O), 6.92 (q, J = 3 Hz et J = 6 Hz, 1H, 5-H);
) d 20.61 (d, J = 1 Hz, 2-Me), 25.68 (d, J = 1 Hz, 4-
–CH –CH ), 30.91 (3-C), 68.48 (d,
), 69.95 (d, J = 8.25 Hz, CH –CH –CH ),
75.37 (d, J = 152.5 Hz, 2-C), 135.04 (d, J = 7.5 Hz, 5-C); Anal. Calcd
for C P: C, 43.84; H, 6.44; N, 6.39. Found: C, 43.83, H,
6.60; N, 6.38.
3
) d 14.72; H NMR (CDCl
3
)
d 11.99 (CHCH
3
2
–CH –CH ),
2
2
C), 30.91 (d, J = 9 Hz, CHCH
3
2
2
2
3
)
2
3
)
2
1H, O–CH
C NMR (CDCl
2
–CH
2
3
–CH
2
1
3
C
9
H18NO
3
P: C, 49.31; H, 8.28; N, 6.39. Found: C, 48.93; H, 8.65;
N, 5.55.
C), 26.48 (d, J = 3 Hz, CH
J = 6.7 Hz, CH –CH –CH
2
2
2
2
2
2
2
2
2
4
.2.2.4.
2-Methyl-2-(5,5-dimethyl-2-oxo-1,3,2-dioxaphosph-
inan-2-yl)pyrrolidine (3c). White crystals. Yield: 80%. Mp
8 4
H14NO
3
1
) d 26.43; 1H NMR (CDCl
) d 1.04 (3H,
3 3
1
10 °C. P NMR (CDCl
CCH ), 1.06 (3H, CCH ), 1.41 (d, J = 15 Hz, 3H, 2-Me), 1.61–1.94
m, 4H, 3-H, 4-H, NH), 2.26–2.38 (m, 1H, 3-H), 2.88–2.96 (m, 1H,
3
3
(
4.2.3.2. trans-2-Methyl-2-(5-methyl-2-oxo-1,3,2-dioxaphosph-
inan-2-yl)-3,4-dihydro-2H-pyrrole 1-oxide[trans-MeCyDEPMPO
1
3
5
(
2
-H), 3.06–3.13 (m, 1H, 5-H), 3.98–4.19 (m, 4H, OCH
CDCl ) d 21.55 (CCH ), 21.72 (CCH ), 24.30 (d, J = 6.8 Hz, 2-Me),
5.80 (d, J = 4.5 Hz, 4-C), 32.75 (d, J = 6.8 Hz, C(CH ), 34.80 (d,
2
); C NMR
31
3
3
3
(trans-4b)]. Yellow crystals. Yield: 81%. Mp 120 °C. P NMR
1
3
)
2
(CDCl
1.74 (d, J = 15 Hz, 3H, 2-CH
(m, 1H, CH–CH ), 2.65–2.76 (m, 2H, 4-H), 2.80–3.05 (m, 1H, 3-H),
4.07–4,14 (m, 1H, OCH ), 4.15–4.30 (m, 2H, OCH ), 4.70–4.78 (m,
1H, OCH ), 6.91 (q, J = 3 Hz et J = 6 Hz, 1H, 5-H); C NMR (CDCl
d 10.89 (CHCH ), 20.65 (d, J = 15 Hz, 2-Me), 25.66 (4-C), 30.84 (3-
C) 30.95 (d, J = 10 Hz, CHCH ), 73.52 (d, J = 6.5 Hz, OCH ), 75.38
(d, J = 150 Hz, 2-C), 75.45 (d, J = 7.5 Hz, OCH ), 134,91 (d,
J = 9.0 Hz, 5-C); Anal. Calcd for C P: C, 46.35; H, 6.92; N,
6.01. Found: C, 46.13; H, 7.13; N, 5.99.
3
) d 13.59; H NMR (CDCl
3 3
) d 0.81 (d, J = 9 Hz, 3H, CH–CH ),
J = 3.8 Hz, 3-C), 47.35 (d, J = 6.8 Hz, 5-C), 60.26 (d, J = 163 Hz, 2-
C), 76.03 (d, J = 7.5 Hz, OCH ), 76.21 (d, J = 7.5 Hz, OCH ); Anal.
Calcd for C10 P: C, 43.52; H, 8.35; N, 7.25. Found: C, 42.37;
H, 8.21; N, 7.39.
3
), 2.16–2.50 (m, 1H, 3-H), 2.48–2.60
2
2
3
H20NO
3
2
2
1
3
2
3
)
3
4
2
.2.2.5. 2-Methyl-2-(5,5-diethyl-2-oxo-1,3,2-dioxaphosphinan-
,
3
2
3
1
-yl)pyrrolidine (3d). White crystals. Yield: 58%. Mp 107 °C.
) d 27.10; 1H NMR (CDCl
) d 0.84 (t, J = 9 Hz, 6H,
), 1.40 (d, J = 15 Hz, 3H, 2-Me), 1.46 (q, J = 9 Hz, 4H,
P
2
NMR (CDCl
CH CH
3
3
9 4
H16NO
2
3

2
228
G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
4
.2.3.3.
cis-2-Methyl-2-(5-methyl-2-oxo-1,3,2-dioxaphosph-
4.3. X-ray diffraction analysis
inan-2-yl)-3,4-dihydro-2H-pyrrole 1-oxide [cis-MeCyDEPMPO
3
1
(
cis-4b)]. Yellow crystals. Yield: 50%. Mp 110 °C. P NMR (CDCl
3
)
Samples of nitrones recrystallized as described above were
washed with cold TBME and dried under vacuum to give single
crystals suitable for X-ray diffraction studies. The intensities were
collected at 293 K on a Bruker-Nonius Kappa CCD diffractometer
1
d 15.53; H NMR (CDCl
J = 15 Hz, 3H, 2-Me), 2.05–2.17 (m, 1H, 3-H), 2.21–2.28 (m, 1H, CH–
CH ), 2.60–2.76 (m, 2H, 4-H), 2.88–3.05 (m, 1H, 3-H), 4.05–4,18 (m,
H, OCH , OCH ), 4.34–4.44 (m, 1H, OCH ), 4.72–4.82 (m, 1H,
OCH ), 6,91 (q, J = 3 Hz, J = 6 Hz, 1H, 5-H); C NMR (CDCl
3.05 (CHCH ), 20.55 (d, J = 15 Hz, 2-Me), 25.61 (4-C), 30.05 (d,
J = 10 Hz, CHCH ), 30.95 (3-C) 73.16 (d, J = 7.5 Hz, OCH ), 73.86
d, J = 7,5 Hz, OCH ), 74.98 (d, J = 153.2 Hz, 2-C), 134,88 (d,
J = 9,8 Hz, 5-C); Anal. Calcd for C P: C, 46.35; H, 6.92; N,
3 3
) d 1.08 (d, J = 12 Hz, 3H, CH–CH ), 1.74 (d,
3
2
2
2
2
using graphite-monochromated Mo K
Kappa CCD program was used for data collection. The crystal
a
radiation (k = 0.71073 Å).
13
53
2
3
) d
5
4
1
3
structures were solved by direct methods and all non-hydrogen
atoms were refined anisotropically by full-matrix least-squares
3
,
2
2
54
(
2
calculations based on F . The structure resolution computing
5
5
9
H16NO
4
was SIR-92.
Crystal data for cis-4b: M = 233.20, triclinic, Pc, a = 6.4636(2) Å,
b = 6.9938(2) Å, c = 12.4669(5) Å, = 87.9420(1)°, b = 77.9340(1)°,
6
.01. Found: C, 45.84; H, 7.19; N, 5.56.
a
3
ꢀ
c
1
= 77.9070(3)°, V = 538.87(3) Å , Z = 2, space group P1, q =
.437 g/cm , l = 2.50 cm , 3873 reflections measured in the
4
.2.3.4.
2-Methyl-2-(5,5-dimethyl-2-oxo-1,3,2-dioxaphosph-
CyDEPMPO
4c)]. Yellow crystals. Yield: 55%. Mp 128 °C. P NMR (CDCl ) d
–O),
–O), 1.76 (d, J = 15 Hz, 3H, 2-Me),
.00–2.14 (m, 1H, 3-H), 2.62–2.74 (m, 2H, 4-H), 2.90–3.03 (m,
H, 3-H), 3.88–3.99 (m, 2H, O–CH –C(CH –CH –O), 4.18–4.23
–CC(CH –CH –O), 4.75–4.80
dd, J = 3 Hz, J = 12 Hz, 1H, O–CH –C(CH –CH –O), 6.91 (q,
) d 20.00 (d, J = 16 Hz,
–O), 25.23 (4-C), 30.37 (3-C),
–CH –O), 74.74 (d, J = 151 Hz, 2-
–C(CH –CH –O), 78.77 (d, J = 7 Hz,
–O), 134.65 (d, J = 8.25 Hz, 5-C); Anal. Calcd
P: C, 48.58; H, 7.34; N, 5.67. Found: C, 48.23; H,
3
À1
inan-2-yl)-3,4-dihydro-2H-pyrrole 1-oxide [Me
2
31
1.67–25.94° h range (1956 independent, Rint = 0.029).
(
3
4,17; 1H NMR (CDCl
Crystal data for 4d: M = 275.28, monoclinic, Pc, a = 6.5915(2) Å,
1
1
2
1
(
(
3
) d 0.89 (s, 3H, O–CH
–C(CH –CH
2
–C(CH
3
2
) –CH
2
3
b = 10.0296(3) Å, c = 20.9770(8) Å, b = 91.161(1)°, V = 1386.51(8) Å ,
.28 (s, 3H, O–CH
2
3
)
2
2
3
À1
Z = 4, space group P2
1
/n, q = 1.319 g/cm , l = 2.05 mm , 11091
reflections measured in the 3.22–28.12° h range (3278 independent,
2
3
)
2
2
Rint = 0.067).
dd, J = 3 Hz et J = 12 Hz, 1H, O–CH
2
3
)
2
2
Crystal data for 4e: M = 275.28, orthorhombic, Pc, a = 12.6103(2) Å,
2
3
)
2
2
3
13
b = 14.0468(3) Å, c = 16.1676(2) Å, V = 2863.84(8) Å , Z = 8, space
J = 3 Hz, J = 6 Hz, 1H, 5-H); C NMR (CDCl
2
3
C), 76.97 (d, J = 7 Hz, O–CH
O–CH –C(CH –CH
for C10
3
3
À1
group Pbca,
q = 1.277 g/cm , l = 1.99 cm , 20423 reflections mea-
2 3 2 2
-Me), 21.50 (O–CH –C(CH ) –CH
sured in the 2.51–25.97° h range (2742 independent, Rint = 0.042).
Supplementary crystallographic data for the structures in this
paper have been deposited with the Cambridge Crystallographic
Data Centre as publication numbers CCDC 793305, 793307,
2.10 (d, J = 9 Hz, O–CH
2
–C(CH
)
3 2
2
2
3
)
2
2
2
3
)
2
2
H18NO
4
7
93315, 793316 and 793317. Copies of the data can be obtained
7
.51; N, 5.53.
free of charge, on application to CCDC, 12 Union Road, Cambridge
CB2 1EZ, UK (fax: +44 1223 336033; E-mail: deposit@ccdc.
cam.ac.uk).
4
.2.3.5. 2-Methyl-2-(5,5-diethyl-2-oxo-1,3,2-dioxaphosphinan-2-
CyDEPMPO (4d)]. Yellow
crystals. Yield: 77%. Mp 96 °C. P NMR (CDCl
) d 15.14; 1H NMR
CDCl ) d 0.78 (t, J = 9 Hz, 3H, CH –CH ), 0.88 (t, J = 6 Hz, 3H,
CH –CH ), 1.22 (q, J = 6 Hz, 2H, CH –CH ), 1.72 (d, J = 15 Hz, 3H,
-Me), 1.74 (q, J = 6 Hz, 2H, CH –CH ), 2.12–2.24 (m, 1H, 3-H),
.60–2.76 (m, 2H, 4-H), 2.90–3.03 (m, 1H, 3-H), 3.98–4.07 (m,
H, O–CH ), 4.18–4.23 (dd, J = 3 Hz et J = 12 Hz, 1H, O–CH ),
.77–4.82 (dd, J = 3 Hz et J = 12 Hz, 1H, O–CH ), 6.91 (q, J = 3 Hz
) d 6.80 (CH CH ), 7.18
), 20.66 (d, J = 15 Hz, 2-Me), 22.07 (CH CH ), 22.64
), 25.69 (4-C), 31.00 (3-C) 37.42 (d, J = 9 Hz, C(CH
–C(CH
yl)-3,4-dihydro-2H-pyrrole 1-oxide [Et
2
31
3
4
.4. Partition coefficient P
(
3
2
3
2
3
2
3
P values were determined using a variation of a previously re-
5
6
2
2
2
4
2
3
ported method. Briefly, 1-octanol-saturated water was obtained
by strongly stirring an equal volume of distilled water and 1-octa-
nol for 1 h at 37 °C, followed by centrifugation (5000 rpm) of the
resulting aqueous layer for 10 min at 10 °C. The resulting homoge-
nous solution was used as a solvent to prepare a stock solution of
each nitrone (typically 20 mg in 10 mL) and to establish a calibra-
tion curve from the measurement of the absorbance at 234 nm of
successive dilutions (up to 0.2 mM). For determination of P values,
a 1-octanol (10 mL)/water (10 mL) solution was prepared and pro-
cessed as described above where 10 mM nitrone was previously
2
2
2
1
3
et J = 6 Hz, 1H, 5-H); C NMR (CDCl
3
2
3
(
(
CH
CH
2
CH
3
2
3
2
CH
3
,
2
CH
3
)
)
2
),
–
7
5.42 (d, J = 149.4 Hz, 2-C), 75.08 (d, J = 7,5 Hz, O–CH
CH –O), 76.88 (d, J = 7.5 Hz, O–CH –C(CH –CH
J = 8.3 Hz, 5-C); Anal. Calcd for C12 P: C, 52.36; H, 8.06; N,
2
3
2
2
2
3
)
2
2
–O), 134.95 (d,
H22NO
4
5
.09. Found: C, 52.24; H, 8.17; N, 5.05.
dissolved in the water portion. A 100-lL aliquot of the recovered
aqueous phase was then diluted 100 times, its absorbance at
234 nm determined and P was calculated from the correponding
calibrationcurve as the ratio of the nitrone concentration in 1-octanol
4
.2.3.6.
2-Methyl-2-(4,4,6,6-tetramethyl-2-oxo-1,3,2-diox-
Cy-
DEPMPO (4e)]. Yellow crystals. Yield: 60%. Mp 156 °C. P NMR
aphosphinan-2-yl)-3,4-dihydro-2H-pyrrole 1-oxide [Me
4
31
to that in water with the following equation: P = [nitrone]1-octanol
/
1
(
CDCl
CH ), 1.53 (d, J = 3 Hz, 3H, C–CH
CH –CCH ), 1.70 (d, J = 15 Hz, 3H, 2-Me), 1.93 (d, J = 12 Hz, 1H,
C(CH –CH –C(CH ), 2,00–2.10 (m, 1H, 3-H), 2.21 (d, J = 12 Hz,
H, C(CH –CH –C(CH ), 2,50–2.56 (m, 1H, 4-H), 2.62–2.74 (m,
H, 4-H), 2.73–3.05 (m, 1H, 3-H), 6.83 (q, J = 3 Hz, J = 6 Hz, 1H, 5-
3
) d 14.64; H NMR (CDCl
3
) d 1.40–1.41 (d, J = 3 Hz, 3H, C–
[nitrone]water. P values are means at least three measurements.
3
3
), 1.60 (d, J = 3 Hz, 6H, CCH
3
–
2
3
4.5. Cell culture and cytotoxicity assays
3
)
2
2
3 2
)
1
1
3
)
2
2
3
)
2
3T3 murine fibroblasts (ATCC-LGC Promochem, Molsheim,
France) were cultured in DMEM supplemented with 1% glucose,
10% fetal calf serum, 100 units/mL penicillin and 100 lg/mL strep-
1
3
H); C NMR (CDCl
0.62 (d, J = 5 Hz, 1 CH
J = 5 Hz, 2CH in C(CH –CH
in C(CH –CH –C(CH ), 31.62 (d, J = 1.5 Hz, 3-C), 46.42 (d,
J = 9 Hz, O–C(CH –CH –C(CH –O), 74.78 (d, J = 155.5 Hz, 2-C),
2.16 (d, J = 7.5 Hz, C(CH ), 82.42 (d, J = 7.5 Hz, C(CH ), 133.94
d, J = 8 Hz, 5-C); Anal. Calcd for C12 P: C, 52.36; H, 8.06;
N, 5.09. Found: C, 51.98; H, 8.22; N, 5.04.
3
) d 20.49 (d, J = 1.5 Hz, 2-Me), 25.27 (4-C),
in C(CH –CH –C(CH ), 30.92 (d,
–C(CH ), 31.40 (d, J = 5 Hz, 1 CH
3
3
3
)
2
2
3
)
2
tomycin at 37 °C in a 5% CO2 incubator. Cells were plated in
3
3
)
2
2
3
)
2
3
24-well dishes and medium was replenished every 2–3 days until
5
7
3
)
2
2
3
)
2
confluency as previously described.
3
)
2
2
3
)
2
In cytotoxicity studies cell necrosis was evaluated by the amount
of lactate dehydrogenase (LDH) release. Aqueous stock solutions
of nitrones (0.3–0.8 M) were freshly prepared and aliquots
were added into each culture well of confluent cells previously
8
(
3
)
2
3 2
)
H22NO
4

G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
2229
filled with phenol red-free DMEM containing 1% glucose, to reach a
final volume of 0.5 mL/well. The final mixtures, which contained
timely monitoring ESR spectra recorded using the KO
2
-driven or
catalase-stopped Fenton systems described above, respectively.
For both types of spin adducts a first-order exponential decay
approximation was performed to determine t1/2app values using
GraphPad Prism software. Data are means of at least three determi-
nations, with 5–10% accuracy.
0
3
.5–100 mM of each tested nitrone, were then incubated for
–24 h (37 °C, 5% CO humidity) and the supernatants were sam-
2
pled for LDH assay using a detection kit (Biolabo, Maizy, France)
according to the manufacturer’s protocol, with light absorption
being measured at 340 nm. Data are expressed as UI/L and are rep-
resentative of 6–12 independent experiments made in duplicate.
Untreated controls represent wells being incubated in DMEM alone
for the same period. After 24 h spin trap exposure, cells were incu-
4
.8. ESR spectroscopy
Throughout the study, the same 10-mm quartz flat cell was
bated for 2 h in serum-free medium containing neutral red (50 lg/
mL). The determination of cell number neutral red uptake was then
used and all procedures allowing to acquire ESR signals under
the strictly controlled time conditions have been described else-
obtained by monitoring the absorption difference between 560 and
32
where. X-band (9.78 GHz) ESR spectra were recorded at room
6
20 nm using a microplate reader.58 Medium lethal dose (LD50
)
temperature with Bruker spectrometers (ESP 300 or EMX) with
was determined as the spin trap concentration inducing 50% cell
death and is expressed in mM from triplicate experiments. When
appropriate, the concentration reducing the viability by 10%
1
00 kHz modulation frequency and 10 mW microwave power.
Other instrument settings in individual experiments are given in
the legends of Figures. Splitting parameters were obtained from
computer simulations of the best resolved signals using the pro-
(
LD10) was determined from the dose–response curve.
5
9
gram of Rockenbauer and Korecz.
4
.6. Statistics
Acknowledgments
Biological values are expressed as mean ± SD. Differences were
analyzed using a one-way ANOVA followed by a a posteriori
Newman–Keuls. Intergroup differences were considered to be
significant at p <0.05.
We thank M. Giorgi for his kind help in collection of crystallo-
graphic data. This study was supported and by the Agence Natio-
nale pour la Recherche (ANR-09-BLAN-0005-03-ROS SIGNAL).
.
4
.7. Free radical generators, HO diastereoisomers assessment
and kinetics of spin adducts
References and notes
Superoxide was produced at room temperature by incubating
the nitrone (0.1–0.15 M) in 20 mM phosphate buffer (pH 7.4)
either with a mixture of 0.4 mM HY, 0.04–0.05 units/mL XO and
1. Harbour, J. R.; Chow, V.; Bolton, J. R. Can. J. Chem. 1974, 52, 3549.
2.
3.
4.
5.
Fréjaville, C.; Karoui, H.; Tuccio, B.; Le Moigne, F.; Culcasi, M.; Pietri, S.;
Lauricella, R.; Tordo, P. J. Med. Chem. 1995, 38, 258.
Olive, G.; Mercier, A.; Le Moigne, F.; Rockenbauer, A.; Tordo, P. Free Radical Biol.
Med. 2000, 28, 403.
Villamena, F. A.; Rockenbauer, A.; Gallucci, J.; Velayutham, M.; Hadad, C. M.;
Zweier, J. L. J. Org. Chem. 2004, 69, 7994.
Karoui, H.; Nsanzumuhire, C.; Le Moigne, F.; Tordo, P. J. Org. Chem. 1999, 64,
1471.
1
(
mM DTPA, or with 18-crown-6 ether (10 mM)-solubilized KO
10 mM) in DMSO.
To generate CyDEPMPOs-OH adducts at room temperature from
2
the parent nitrone (0.05–0.1 M) four methods were alternatively
used: (i) reduction of preformed CyDEPMPO-OOH in 20 mM phos-
phate buffer (as described above) by a mixture of GPx and GSH, as
6.
7.
8.
Stolze, K.; Udilova, N.; Nohl, H. Free Radical Biol. Med. 2000, 29, 1005.
Chalier, F.; Tordo, P. J. Chem. Soc., Perkin Trans. 2 2002, 2110.
Xu, Y. K.; Chen, Z. W.; Sun, J.; Liu, K.; Chen, W.; Shi, W.; Wang, H. M.; Liu, Y. J.
Org. Chem. 2002, 67, 7624.
3
2
detailed in; (ii) incubation with a Fenton reagent consisting of
mM FeSO and 2 mM H in 20 mM phosphate buffer; (iii) pho-
tolysis 1% H in degassed phosphate buffer; (iv) incubation with
mM FeCl in water.
The CyDEPMPOs adducts of CH , CH
9.
Hardy, M.; Chalier, F.; Finet, J. P.; Rockenbauer, A.; Tordo, P. J. Org. Chem. 2005,
1
4
2 2
O
7
0, 2135.
2 2
O
10. Kamibayashi, M.; Oowada, S.; Kameda, H.; Okada, T.; Inanami, O.; Ohta, S.;
1
3
Ozawa, T.; Makino, K.; Kotake, Y. Free Radical Res. 2006, 40, 1166.
11. Hardy, M.; Rockenbauer, A.; Vásquez-Vivar, J.; Felix, C.; Lopez, M.; Srinivasan,
Å
.
Å
ÅÀ
3
CHOH, GS and CO were
2
3
S.; Avadhani, N.; Tordo, P.; Kalyanaraman, B. Chem. Res. Toxicol. 2007, 20, 1053.
produced in 20 mM phosphate buffer by incubating the parent nit-
rone (0.05 M) with a mixture of the Fenton reagent described
above and the precursor (0.05–0.2 M) DMSO, ethanol, GSH and
12. Chalier, F.; Hardy, M.; Ouari, O.; Rockenbauer, A.; Tordo, P. J. Org. Chem. 2007,
7
2, 7886.
13. Hardy, M.; Bardelang, D.; Karoui, H.; Rockenbauer, A.; Finet, J. P.; Jicsinszky, L.;
Rosas, R.; Ouari, O.; Tordo, P. Chem. Eur. J. 2009, 15, 11114.
NaHCO
ial oxidation of a mixture of aqueous nitrone (0.1 M) and ꢀ5 mg of
NaBH . CyDEPMPOs-OtBu and CyDEPMPOs-OOtBu adducts were
obtained in degassed anhydrous toluene by continuous UV photol-
2
, respectively. CyDEPMPOs-H adduct was prepared by aer-
14. Zhao, H.; Joseph, J.; Zhang, H.; Karoui, H.; Kalyanaraman, B. Free Radical Biol.
Med. 2001, 31, 599.
4
15. 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.
1
6. Stolze, K.; Udilova, N.; Rosenau, T.; Hofinger, A.; Nohl, H. Biol. Chem. 2003, 384,
93.
17. Stolze, K.; Udilova, N.; Rosenau, T.; Hofinger, A.; Nohl, H. Biochem. Pharmacol.
005, 69, 297.
2
ysis of the parent nitrone (0.05 M) with 10 mM (tBuO) or tBuOOH,
respectively.
4
2
The cis/trans ratios of CyDEPMPOs-OH adducts were deter-
Å
18. Stolze, K.; Rohr-Udilova, N.; Rosenau, T.; Stadtmüller, R.; Nohl, H. Biochem.
Pharmacol. 2005, 69, 1351.
19. Stolze, K.; Rohr-Udilova, N.; Rosenau, T.; Hofinger, A.; Kolarich, D.; Nohl, H.
Bioorg. Med. Chem. 2006, 14, 3368.
20. Stolze, K.; Rohr-Udilova, N.; Rosenau, T.; Hofinger, A.; Nohl, H. Bioorg. Med.
Chem. 2007, 15, 2827.
21. Xu, Y.; Kalyanaraman, B. Free Radical Res. 2007, 41, 1.
22. Stolze, K.; Rohr-Udilova, N.; Hofinger, A.; Rosenau, T. Bioorg. Med. Chem. 2008,
mined for each of the four HO generators described above by com-
puter analyzing the ESR spectra obtained immediately after spin
adduct formation was considered over, that is, after cut-off of the
UV lamp or addition of Fe3 in the case of H
+
2
O
2
photolysis or nucle-
ophilic addition of water, respectively, or after complete removal of
ÅÀ
the spin adduct precursor, that is, O2 (by excess SOD+catalase) and
1
6, 8082.
3. Stolze, K.; Rohr-Udilova, N.; Hofinger, A.; Rosenau, T. Bioorg. Med. Chem. 2009,
7, 7572.
24. Buettner, G. R.; Oberley, L. W. Biochem. Biophys. Res. Commun. 1978, 83, 69.
2 2
H O (by excess catalase alone) for the reduction of CyDEPMPOs-
OOH or the Fenton reaction, respectively. To minimize simulation
errors due to fitting of the linewidths the hfcs reported in Table
2
1
2
2
2
5. Shi, H.; Timmins, G.; Monske, M.; Burdick, A.; Kalyanaraman, B.; Liu, Y.;
Clément, J. L.; Burchiel, S.; Liu, K. J. Arch. Biochem. Biophys. 2005, 437, 59.
6. Chen, Y. R.; Chen, C. L.; Yeh, A.; Liu, X.; Zweier, J. L. J. Biol. Chem. 2006, 281,
13159.
4
for diastereoisomers of CyDEPMPOs-OH were set as non-variable
parameters.
The kinetic parameters of CyDEPMPOs-OOH and CyDEPMPOs-
OH adducts were extracted from decay curves constructed by
7. Karoui, H.; Rockenbauer, A.; Pietri, S.; Tordo, P. Chem. Commun. 2002, 3030.

2
2
230
G. Gosset et al. / Bioorg. Med. Chem. 19 (2011) 2218–2230
8. Bardelang, D.; Rockenbauer, A.; Karoui, H.; Finet, J. P.; Biskupska, I.; Banaszak,
K.; Tordo, P. Org. Biomol. Chem. 2006, 4, 2874.
43. Pietri, S.; Liebgott, T.; Fréjaville, C.; Tordo, P.; Culcasi, M. Eur. J. Biochem. 1998,
254, 256.
2
3
9. Han, Y.; Tuccio, B.; Lauricella, R.; Villamena, F. A. J. Org. Chem. 2008, 73, 7108.
0. Han, Y.; Liu, Y.; Rockenbauer, A.; Zweier, J. L.; Durand, G.; Villamena, F. A. J. Org.
Chem. 2009, 74, 5369.
1. Makino, K.; Hagiwara, T.; Hagi, A.; Nishi, M.; Murakami, A. Biochem. Biophys.
Res. Commun. 1990, 172, 1073.
2. Culcasi, M.; Rockenbauer, A.; Mercier, A.; Clément, J. L.; Pietri, S. Free Radical
Biol. Med. 2006, 40, 1524.
3. Page, P.; Mazières, M. R.; Bellan, J.; Sanchez, M. Phosphorus, Sulfur Silicon Relat.
Elem. 1992, 70, 205.
44. Cruz-Gregorio, S.; Rodriguez-Palacios, V.; Höpfl, H.; Quintero, L.; Sartillo-Piscil,
F. J. Org. Chem. 2009, 74, 197.
45. Clément, J. L.; Finet, J. P.; Fréjaville, C.; Tordo, P. Org. Biomol. Chem. 2003, 1,
1591.
46. Rockenbauer, A.; Clément, J. L.; Culcasi, M.; Mercier, A.; Tordo, P.; Pietri, S. J.
Phys. Chem. A 2007, 111, 4950.
47. Karoui, H.; Hogg, N.; Fréjaville, C.; Tordo, P.; Kalyanaraaman, B. J. Biol. Chem.
1996, 271, 6000.
48. Morris, C. J.; Earl, J. R.; Trenam, C. W.; Blake, D. R. Int. J. Cell Biol. 1995, 27, 109.
49. Halliwell, B.; Gutteridge, J. M. C. Arch. Biochem. Biophys. 1986, 246, 501.
50. Klosinski, P. Tetrahedron Lett. 1990, 31, 2025.
51. Maffei, M.; Buono, G. Tetrahedron 2003, 59, 8821.
52. Satish Kumar, N. S.; Kumaraswamy, S.; Said, M. A.; Kumara Swamy, K. C. Org.
Process Res. Dev. 2003, 7, 925.
3
3
3
3
3
4. Mosbo, J. A.; Verkade, J. G. J. Am. Chem. Soc. 1973, 95, 204.
5. Barbati, S.; Clément, J. L.; Fréjaville, C.; Bouteiller, J. C.; Michel, J. C.; Yadan, J. C.;
Tordo, P. Synthesis 1999, 2036.
3
6. Murahashi, S. I.; Mitsui, H.; Shiota, T.; Tsuda, T.; Watanabe, S. J. Org. Chem.
1
990, 55, 1736.
3
3
7. Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
8. Cova, D.; De Angelis, L.; Monti, E.; Piccinini, F. Free Radical Res. Commun. 1992,
53. Nonius, B. V. Kappa CCD Reference Manual; PO Box 811, 2600 Av. Delft, The
Netherlands, 1998.
1
5, 353.
54. Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, 64, 112.
55. Almatore, A.; Cascarano, G.; Giocovazzo, C.; Guagliardi, A.; Burla, M. C.;
Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27, 435.
56. Janzen, E. G.; West, M. S.; Kotake, Y.; DuBose, C. M. J. Biochem. Biophys. Methods
1996, 32, 183.
3
4
4
4
9. Haseloff, R. F.; Mertsch, K.; Rohde, E.; Baeger, I.; Grigor’ev, I. A.; Blasig, I. E. FEBS
Lett. 1997, 418, 73.
0. Arakaki, N.; Kajihara, T.; Arakaki, R.; Ohnishi, T.; Kazi, J. A.; Nakashima, H.;
Daikuhara, Y. J. Biol. Chem. 1999, 274, 13541.
1. Rohr-Udilova, N.; Stolze, K.; Marian, B.; Nohl, H. Bioorg. Med. Chem. Lett. 2006,
57. Culcasi, M.; Muller, A.; Mercier, A.; Clément, J. L.; Payet, O.; Rockenbauer, A.;
Marchand, V.; Pietri, S. Chem. Biol. Interact. 2006, 164, 215.
58. Borenfreund, E.; Puerner, J. A. Toxicol. Lett. 1985, 24, 119.
59. Rockenbauer, A.; Korecz, L. Appl. Magn. Reson. 1996, 10, 29.
16, 541.
2. Rohr-Udilova, N.; Stolze, K.; Sagmeister, S.; Parzefall, W.; Marian, B.; Nohl, H.;
Schulte-Hermann, R.; Grasl-Kraupp, B. Bioorg. Med. Chem. Lett. 2007, 17, 5698.