Synthesis and properties of a series of β-cyclodextrin/nitrone spin traps for improved superoxide detection

Article abstract of DOI:10.1039/c7ob00961e

Three new DEPMPO-based spin traps have been designed and synthesized for improved superoxide detection, each carrying a cyclodextrin (CD) moiety but with a different alkyl chain on the phosphorus atom or with a long spacer arm. EPR spectroscopy allowed us to estimate the half-life of the superoxide spin adducts which is close to the value previously reported for CD-DEPMPO (t_1/2 ≈ 50-55 min under the conditions investigated). The spectra are typical of superoxide adducts (almost no features of the HO adduct that usually forms with time for other nitrone spin traps such as DMPO) and we show that at 250 μM, the new spin trap enables the reliable detection of superoxide by 1 scan at the position opposite to the corresponding spin trap without the CD moiety. The resistance of the spin adducts to a reduction process has been evaluated, and the superoxide spin adducts are sensitive to ascorbate and glutathione (GSH), but not to glutathione peroxidase/GSH, reflecting the exposed nature of the nitroxide moiety to the bulk solvent. To understand these results, 2D-ROESY NMR studies and molecular dynamics pointed to a shallow or surface self-inclusion of the nitrone spin traps and of nitroxide spin adducts presumably due to the high flexibility of the permethyl-β-CD rim.

Full text of DOI:10.1039/c7ob00961e

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Synthesis and properties of a series of
β-cyclodextrin/nitrone spin traps for
improved superoxide detection†
Cite this: DOI: 10.1039/c7ob00961e
a
a
b
c
Florent Poulhès, Egon Rizzato, Patrick Bernasconi, Roselyne Rosas,
d,e
f
g
a
Stéphane Viel,
Didier Siri,
Laszlo Jicsinszky, Antal Rockenbauer, David Bardelang,
h
h
a
a
Anouk Gaudel-Siri, Hakim Karoui, Micaël Hardy* and
a
Olivier Ouari
*
Three new DEPMPO-based spin traps have been designed and synthesized for improved superoxide
detection, each carrying a cyclodextrin (CD) moiety but with a different alkyl chain on the phosphorus
atom or with a long spacer arm. EPR spectroscopy allowed us to estimate the half-life of the superoxide
spin adducts which is close to the value previously reported for CD-DEPMPO (t1/2 ≈ 50–55 min under the
•
conditions investigated). The spectra are typical of superoxide adducts (almost no features of the HO
adduct that usually forms with time for other nitrone spin traps such as DMPO) and we show that at
250 μM, the new spin trap enables the reliable detection of superoxide by 1 scan at the position opposite
to the corresponding spin trap without the CD moiety. The resistance of the spin adducts to a reduction
process has been evaluated, and the superoxide spin adducts are sensitive to ascorbate and glutathione
(GSH), but not to glutathione peroxidase/GSH, reflecting the exposed nature of the nitroxide moiety to
the bulk solvent. To understand these results, 2D-ROESY NMR studies and molecular dynamics pointed
to a shallow or surface self-inclusion of the nitrone spin traps and of nitroxide spin adducts presumably
due to the high flexibility of the permethyl-β-CD rim.
Received 19th April 2017,
Accepted 6th July 2017
DOI: 10.1039/c7ob00961e
rsc.li/obc
radicals with a diamagnetic probe (nitroso or nitrone spin trap)
to form a more persistent nitroxide spin adduct that can
Introduction
Although EPR spectroscopy is the method of choice to study accumulate, and subsequently be detected and characterized by
paramagnetic species, only a few free radicals, such as ascorbyl, EPR. Over the past few decades, the EPR spin trapping tech-
trityl and nitroxide radicals, exhibit long-lived features allowing nique has evolved as one of the gold standard methods to inves-
1–5
their direct detection by EPR under physiological conditions. tigate transient free radicals. While spin trapping can provide
One of the approaches to detect transient free radicals by EPR detailed insights into the structure, kinetics of generation or
is spin trapping. This technique relies on the reaction of free decay and concentration of free radicals in organic solvents or
in buffer solutions, its application to biological samples is still
challenging for the study of Reactive Oxygen Species (ROS),
a
b
c
mainly due to the slow kinetics of superoxide trapping and the
rapid decay of the generated spin adducts into diamagnetic by-
Aix Marseille Univ, CNRS, ICR, UMR 7273, SREP, 13013 Marseille, France
Aix Marseille Univ, CNRS, ICR, UMR 7273, 13013 Marseille, France
Aix Marseille Univ, CNRS, Spectropole, 13013 Marseille, France
products by their reaction with bioconstituents, such as peroxi-
d
6–8
Aix Marseille Univ, CNRS, ICR, UMR 7273, SACS, 13013 Marseille, France
Institut Universitaire de France, 75005 Paris, France
dase enzymes, ascorbate, and thiols.
Due to the important
e
f
9
but complex biological role of ROS as signaling molecules, and
in pathologic situations such as atherosclerosis, neurodegenera-
tive diseases, diabetes, and cancer, there is a crucial need to
develop methods or techniques to investigate ROS production
in biological systems. Studies focused on improving the intrin-
sic properties of the spin traps have received a lot of attention,
and in the past two decades new generations of spin traps have
been developed exhibiting long-lived spin adducts, targeting
University of Turin, Dipartimento di Scienza e Tecnologia del Farmaco,
via P. Giuria 9, 10125 Torino, Italy
g
Institute of Materials and Environmental Chemistry, Res. Centre for Natural
Sciences, Hungarian Academy of Sciences, Budapest, Hungary. Dpt of Physics,
Budapest University of Technology and Economics and MTA-BME, PROSPIN,
P.O. Box 91, H-1521 Budapest, Hungary
h
Aix Marseille Univ, CNRS, ICR, UMR 7273, CT, 13013 Marseille, France
†
Electronic supplementary information (ESI) available: Details of syntheses,
NMR spectra of the prepared compounds and details of the MD simulations. See
DOI: 10.1039/c7ob00961e
1
0–20
features and/or faster superoxide trapping rates.
However,
This journal is © The Royal Society of Chemistry 2017
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
the bioreduction of the spin adducts is still a bottleneck for with the aim to better understand the key parameters for
spin trapping experiments in biological systems. Advances such improving their spin trapping performance.
2
1,22
as rapid-scan EPR
and spin trapping combined with
Using a 4-hydroxymethyl-DEPMPO derivative as a spin trap
2
3
immunological detection (immunospin trapping) have been building block, we evaluated the effects of modifying the
developed in the recent years, but there is still a need for spin spacer arm, and the bulkiness and the length of the alkyl
adducts more resistant to bioreduction. One promising strategy chains of the phosphonate group on the spin trapping per-
relies on the selective inclusion of the generated spin adducts formance and on the positioning of the spin trap and spin
inside a macrocyclic cavity such as cyclodextrins (CDs) to phys- adduct moieties relative to the CD cavity. Combining EPR and
ically remove the reduction sensitive nitroxide moiety from the NMR studies and molecular modelling, the nitrone group was
bulk or the bioconstituents. Although the concept of protecting estimated to be mainly lying near the CD primary rim. The life-
a nitroxide by complexation in CD has been reported and time of the superoxide adducts and the protection towards
2
4–26
studied by several authors since the 70s,
its transposition to GPX have shown to improve for the three new spin traps com-
spin trapping has appeared two decades later, using non- pared with those of DEPMPO or 5-(diisopropoxyphosphoryl)-5-
27–35
covalent and covalent approaches.
In the past few years, methyl-1-pyrroline-N-oxide (DIPPMPO) alone, with the best
this method has been used successfully in various systems such results obtained for CD-BEPMPO. The new molecules enabled
36
37
as cell cultures, fuel cell membranes and thylakoids. For the detection of superoxide radicals at 250 μM spin trap con-
instance, using 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) or centration, well-below the DIPPMPO concentration for a
(
(
(
2-methyl-3,4-dihydro-1-oxide-2H-pyrrol-2-yl)diethylphosphonate similar reaction.
DEPMPO) as spin traps in the presence of methylated β-CD
6 equivalents), we observed a seven-fold increase of the super-
oxide adduct lifetime and more noticeably a partial protection
2
9
from reduction by thiols and ascorbate. When using a CD- Results and discussion
linked DEPMPO spin trap, similar improvements were observed,
Synthesis
with an enhanced lifetime of the spin adducts in the presence
31,38,39
CD-BEPMPO 9 was prepared (16% overall yield) in a five-step
sequence (Scheme 2) following the procedure described by
Hardy et al. (ESI†). Nitrophosphonate 4 was obtained by a
of microsomes or cells.
Surprisingly, only a few studies
have been devoted to the rational design of such CD spin trap
conjugates to establish the fine relationship between the struc-
tural changes and the spin trapping properties. Several ques-
tions regarding the protecting effects, the optimal linker length
1
7
42
4
step synthesis by adapting the procedure of Hoffmann et al.
Briefly, nitrofuranone 5 was obtained in 80% yield by reacting
nitrophosphonate 4 with 2-(5H)-furanone in the presence of
tributylphosphine as a catalyst. The reduction of 5 by DIBAL-H
at −78 °C led to hemiacetal 6 in good yield (55%). The reduc-
3
1,33,40,41
and the prepositioning of the spin trap remain open.
In
the present work, we report the synthesis, EPR, NMR and model-
ling studies of a series of 3 new CD-linked spin traps (Scheme 1)
Scheme 2 Reagents and conditions: (i) PBu
temperature (rt); (ii) DIBAL-H, CH Cl , −78 °C; (iii) Zn/NH
rt; (iv) DSC, Et N, CH CN, rt; and (v) permethylated 6-monoamino
β-cyclodextrin, Et N, CH Cl , rt.
3
, C
6
H
12/CH
2
Cl
2
, room
2
2
4
Cl, H
2
O/THF,
3
3
Scheme 1 Chemical structures of the investigated spin traps.
3
2
2
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Organic & Biomolecular Chemistry
Paper
tive cyclization of nitroaldehyde 6 in the presence of zinc and lated EPR spectra, obtained using the parameters reported in
ammonium chloride afforded nitrones 7 and 7′ as a mixture of Table 1, are shown as gray lines in Fig. 1. Moreover, when the
cis/trans diastereoisomers, which were separated by silicagel KO /18-crown-6-ether/DMSO system is used, the same pattern
2
column chromatography (59% yield for 7). The reaction of the of the spectra was obtained. In the presence of superoxide dis-
cis diastereoisomer with N,N′-disuccinimidylcarbonate (DSC) mutase (SOD), no EPR signal was detected. The observed EPR
in the presence of triethylamine afforded NHS-BEPMPO 8 in signals were assigned to the corresponding superoxide spin
9
5% yield. Compounds 9–11 were obtained by reacting adducts. All spectra appear as doublets of distorted quartets
1
7
16
31
14
1
NHS-BEPMPO, NHS-DIPPMPO or NHS-DEPMPO with the resulting from hyperfine couplings with P, N and Hβ
appropriate CD, with yields ranging from 58 to 80%.
atoms (Table 1). The spectra were calculated using the EPR/
4
3
ROKI program. For all the series, the best fits were obtained
assuming both that the trapping reaction yields only the trans-
diastereoisomer (hydroperoxyl moiety on the opposite face to
phosphonate and CD) and a chemical exchange occurs
EPR/spin trapping
Fig. 1a–d show the EPR spectra obtained after the incubation
of a mixture containing hypoxanthine (HX) (0.4 mM), xanthine
between two exchanging conformational sites T and T . In the
−
1
1
2
oxidase (XO) (0.04 U mL ), DTPA (1 mM) and the spin trap
20 mM, CD-DIPPMPO, CD -DEPMPO, CD-BEPMPO and
case of the DEPMPO-OOH and DIPPMPO-OOH spin adducts,
the same assumptions were used to account for the large alter-
nated line-width observed.
(
6
DIPPMPO) in phosphate buffer (0.1 M, pH 7.3). In this work,
spin trapping experiments using DIPPMPO were performed
under the same conditions as a reference. DIPPMPO and
DEPMPO exhibit very close spin trapping properties. The simu-
Hydroxyl-adducts for spin traps 9–11
The twelve-line EPR spectra of the hydroxyl adducts of
compounds 9–11 (Fig. 1 right, e–h) were obtained by the
reduction of the corresponding superoxide adducts by the
glutathione peroxidase/glutathione (GPX/GSH) system (Table 1).
Generating hydroxyl radicals by the Fenton reaction led to
complex EPR spectra due to the side-reactions with the CD
moiety.
Decay kinetics of superoxide adducts
The kinetic studies were performed at 296 K in 50 μL glass
capillaries (ESI†). Superoxide was generated enzymatically in
the presence of the studied nitrones. Once the concentration
of the superoxide spin adduct reached a plateau (after
∼
10–14 min), its formation was stopped by the addition of
−
1
SOD (600 U mL ). The signal decay was monitored during the
0 min recording of successive spectra every 42 s (100 spectra)
Fig. S2†).
The decay curves for the superoxide spin adducts of 9–11
7
(
and DIPPMPO were obtained by simulating simultaneously all
the spectra using the ROKI program (Fig. 2). The apparent
half-life times were then determined and are listed in Table 1.
Longer lifetimes are obtained for all the CD-linked spin traps
compared to that of DIPPMPO and the values are roughly
6
similar for the three molecules (CD-DIPPMPO, CD -DEPMPO,
and CD-BEPMPO), indicating a small effect of the structural
change on the superoxide adduct persistence.
Fig. 1 Spin trapping of superoxide radicals using CD-DIPPMPO 11,
6
CD -DEPMPO 10, and CD-BEPMPO 9. (a) EPR spectrum obtained after
1
4 min incubation of a mixture containing hypoxanthine (HX) (0.4 mM),
−1
xanthine oxidase (XO) (0.04 U mL ), DTPA (1 mM) and CD-DIPPMPO Evaluation of the limit of detection
(
20 mM) in oxygen-bubbled phosphate buffer (0.1 M, pH 7.3). (b) As in
A series of experiments at lower spin trap concentrations were
(
a) but containing CD -DEPMPO. (c) As in (a) but containing
6
performed to evaluate the threshold to detect superoxide rad-
CD-BEPMPO (20 mM). (d) As in (a) but containing DIPPMPO. (e) As in (a)
but in the presence of GSH (1 mM) after 20 min. (f) As in (b) but in the icals generated with a flux of 4.2 μM min− using the HX/XO
1
presence of GSH (1 mM) after 20 min. (g) As in (c) but in the presence of
GSH (1 mM) after 20 min. (h) As in (d) but in the presence of GSH (1 mM)
after 20 min. Grey lines: calculated spectra (Table 1). Spectrometer set-
tings: microwave power, 10 mW (a–h); modulation amplitude, 0.7 G (a–
h); smooth point, 1 (a–h); gain, 75 dB (a–h); sweep time, 41.94 s (a–h);
conversion time, 40.96 ms (a–h).
system. At 250 μM spin trap concentration, and after
11–16 min of incubation, the EPR signals of the corresponding
superoxide adducts were observed with a reasonable signal-to-
noise ratio for compounds 9, 10 and 11, whereas a very weak
EPR signal was obtained for DIPPMPO (Fig. 3).
This journal is © The Royal Society of Chemistry 2017
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Calculated EPR parameters for the superoxide and hydroxyl spin adducts. Exchange rate constants in s−1
k (s−
1
)
<a
N
> (G)
<aHβ> (G)
t1/2 (min)
a
Spin adduct
Diastereoisomer
Site
P
<a > (G)
8
8
8
8
8
CD-DEPMPO-OOH
trans (100%)
T
T
T
T
T
T
T
T
T
T
1
2
1
2
1
2
1
2
1
2
(70%)
(30%)
(60%)
(40%)
(88%)
(12%)
(54%)
(46%)
(57%)
(53%)
0.1 × 10
5.1 × 10
4.2 × 10
2.5 × 10
0.2 × 10
53.27
52.01
54.03
51.86
53.41
48.05
53.07
45.01
49.13
51.02
42.09
48.6
12.79
12.97
12.80
12.82
12.86
12.76
12.83
12.69
13.28
12.93
13.83
13.0
12.37
10.13
12.78
10.57
12.13
7.38
12.17
6.06
10.23
12.34
9.03
35 (ref. 31)
56 (2)
CD-BEPMPO-OOH (9-OOH)
trans (100%)
trans (100%)
trans (100%)
trans (90%)
6
CD -DEPMPO-OOH (10-OOH)
54 (1.9)
48 (1.7)
28
CD-DIPPMPO-OOH (11-OOH)
DIPPMPO
cis (10%)
trans (90%)
8
DEPMPO
T
T
1
(50%)
(50%)
0.8 × 10
10.0
12.0
17 (ref. 31)
2
51.2
13.1
cis (10%)
40.6
47.1
13.4
14
9.0
13
CD-DEPMPO-OH
9
1
1
-OH
0-OH
1-OH
—
—
—
—
—
—
—
—
—
52.90
53.02
52.99
13.50
13.52
13.56
10.20
10.29
10.46
a
Apparent half-lifetime values for superoxide adducts of 9–11, and the ratio values of t1/2 9–11-OOH/t1/2 DIPPMPO-OOH in parentheses.
Decay kinetics of superoxide adducts in the presence of
ascorbate, glutathione or GSH/GPX
The protective effect of the CD moiety was evaluated by adding
GPX/GSH, GSH, and ascorbate as the reducing agents. Once
the concentration of the superoxide spin adduct reached a
plateau (10–14 min), SOD was added and then the mixture of
GSH/GPX (0.5 mM and 10 U mL− respectively) acted as
the reducing system converting the hydroperoxyl groups to
hydroxyl groups. In the present series of CD-spin-traps
investigated, no transformation of the adduct-OOH was
observed (Fig. 4), whereas DIPPMPO-OOH was converted to
DIPPMPO-OH adduct (Fig. 4d). When GSH alone was used at a
1
Fig. 2 Kinetics of the decay of CD-DIPPMPO-OOH (11-OOH), CD
DEPMPO-OOH (10-OOH), CD-BEPMPO-OOH (9-OOH) and
DIPPMPO-OOH. Kinetics of the decay of the spin adducts generated by
6
-
concentration of ca. 1 mM, the conversion of adduct-OOH to
adduct-OH was observed, indicating an improved but limited
HX (0.4 mM), XO (0.04 U mL− ), and DTPA (1 mM) in an oxygenated protection relative to DIPPMPO (Fig. S3†).
1
A complete
phosphate buffer (0.1 M, pH 7.4), and the addition of SOD (600 U mL− ).
1
reduction of the EPR signal was obtained in the presence of
ascorbate (500 μM). From these results, it appears that the pro-
tective effect is related to the bulkiness of the CD-spin
adducts, hampering the access to the GPX enzyme active site.
In order to gain details on the actual role of the closed CD in
the protection of the spin adducts, NMR and molecular model-
ling were performed (see below).
Liquid-state NMR study
In our previous work on CD-DEPMPO, the pyrrolidine-N-oxide
ring was estimated to be located above the methoxy crown of
3
1
the primary rim. For the reported new spin traps, i.e.
CD-DIPPMPO, CD -DEPMPO or CD-BEPMPO, more hydro-
6
phobic groups (isopropyl or butyl) or longer spacer arms were
introduced on the spin trap, leading to a possible larger pro-
pensity of the spin trap to enter the cavity. 2D NMR experi-
Fig. 3 Detection of superoxide radicals using 250 μM of nitrone spin
traps and generated by HX (0.4 mM), XO (0.04 U mL ), and DTPA
−1
ments were performed in D O at 20 mM.
2
(
1 mM) in an oxygenated phosphate buffer (0.1 M, pH 7.4). Spectrometer
The 2D ROESY spectrum of CD-DIPPMPO showed intense
cross-peaks between the H-atoms of the nitrone moiety and
the primary methoxy groups, indicating contacts between the
settings: microwave power, 10 mW; modulation amplitude, 0.7 G; time
5
constant, 0.128 s; gain, 10 ; sweep time, 41.9 s; conversion time,
40.9 ms, 1 scan.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Organic & Biomolecular Chemistry
Paper
−1
Fig. 4 (a) Decay kinetics of superoxide spin adducts generated by HX (0.4 mM), XO (0.04 U mL ), and DTPA (1 mM) in an oxygenated phosphate
buffer (0.1 M, pH 7.4), and the addition of SOD (600 U mL ), GPX (10 U mL−1) and GSH (0.5 mM). (a, e) EPR signals of CD-DIPPMPO-OOH after
−1
1
min and 1 h 15 min respectively. (b, f) EPR signals of CD -DIPPMPO-OOH after 1 min and 1 h 15 min respectively. (c, g) EPR signals of
6
CD-BEPMPO-OOH after 1 min and 1 h 15 min respectively. (d) EPR signal of DIPPMPO-OH after 1 min.
spin trap and the primary methoxy crown (Fig. 5a, pastel red). (Fig. 5b) showed evidence for both intra- (pastel red) and inter-
The chemical shift values of the inner cavity protons H(3) molecular (pastel yellow) types of association. No cross-peaks
(
3.72–3.77 ppm) and H(5) (3.86–3.95 ppm) were evaluated by were observed above 3.7 ppm (H(3) and H(5) from the inner
HSQC experiments. The 2D ROESY spectra (Fig. 5a) are in line cavity) indicating no deep inclusion of the spin trap ring.
with the space interactions between the methyl groups of the The absence of the previously observed correlation between
diisopropyl groups of the spin trap and the methoxy groups of the methyl group on the pyrrolidine ring and the primary
the permethylated β-CD derivative. Since these cross-peaks are methoxy crown supports a positioning further apart from the
all situated below 3.7 ppm, the formation of a deeply included primary methoxy crown. These data support the occurrence of
self-inclusion complex involving H(3) and H(5) is not expected. intermolecular type complexes, but the strength of the associ-
Moreover, the methyl group of the pyrrolidine ring exhibits a ation mode is weak as no significant complexation induced
strong correlation with the primary methoxy groups (Fig. 5a, shifts (CIS) can be measured in the range of concentrations
pastel red, left). As for CD-DEPMPO, there is a weak correlation investigated. The DOSY experiments confirm these results
between the isopropyl groups and H(6)CD. Unexpectedly, two (ESI†).
additional cross-peaks involved the methoxy groups on the CD
For CD-BEPMPO, an asymmetric phosphonate group was
(O(2)Me and O(3)Me in Fig. 5a, excerpt, pastel yellow), posi- introduced carrying a linear hydrophobic butyl chain for com-
tioned at the secondary rim (opposite side compared to the parison with CD-DIPPMPO. Strong correlations are observed
nitrone grafting site that is on the primary rim). These cross- between the primary methoxy crown (O(6)Me) and (i) methyl
peaks indicate a possible inter-molecular interaction. These and methylene protons belonging to the two alkyl groups of
results were confirmed by diffusion NMR (DOSY) experiments the phosphonate and (ii) the methyl group of the pyrrolidine
showing the presence of dimer type assemblies (ESI†). All ROE ring (Fig. 5c, pastel red). The intensity of the cross-peaks is
data were confirmed by 1D selective experiments (mixing stronger, suggesting a more prominent nesting of the spin trap
times of 250 and 400 ms) in which the signal of the methyl ring in the surrounding provided by the primary methoxy
6
groups of the diisopropoxyphosphoryl unit was selectively irra- crown (OMe ). This is further supported by the appearance of
diated and the changes in the protons of the CD region were two additional weak correlations between the methyl group at
carefully monitored. The moderate values of complexation the end of the ethoxy and butoxy chains and inner cavity
induced shifts (CIS) as a function of the CD-DIPPMPO concen- protons H(5) or H(6a) (Fig. 5c, pastel yellow, bottom).
tration (CIS ≤ 0.06 ppm) suggests that the inclusion of the
spin trap ring is poorly significant.
Thus, the introduction of a longer hydrophobic group
(ethoxy or butoxy) compared with CD-DIPPMPO (diisopropoxy)
6
For CD -DEPMPO, NMR studies were also performed to in terms of spheric/steric congestion significantly improves the
evaluate the effect of a longer and flexible spacer arm (6 accommodation of the nitrone core in the CD cavity primary
additional atoms) on the inclusion process. This change rim and allows a slightly better self-inclusion (butoxy end).
6
allows for a direct comparison with CD-DEPMPO since only However, as for CD -DEPMPO and CD-DIPPMPO, some inter-
the length of the linker changed. The 2D ROESY spectrum molecular associations are also possible, as evidenced by the
This journal is © The Royal Society of Chemistry 2017
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
cross-peaks involving the methoxy groups at the secondary rim
(
O(2)Me, O(3)Me in Fig. 5c, pastel yellow).
In summary, the longer arm of CD
6
-DEPMPO did not
improve the self-positioning near the entrance of the CD.
Introducing more hydrophobic groups (CD-DIPPMPO and
CD-BEPMPO) yielded a modest change with a positioning
slightly closer to the CD entrance. Modelling studies have
been performed to get further structural and dynamic
information.
Molecular dynamics (MD) of the CD-spin traps (see the ESI†
for details)
MD was carried out for a 50 ns period in a water box as a
means to determine the position of the spin trap with respect
to the CD cavity or the primary methoxy rim. A first obser-
vation is the high flexibility of the CD macrocycle and the
tilting of some glucopyranose units so that one methoxy
group is positioned near the center of the cavity (red arrows,
Fig. 6).
Thus the barycenter was defined according to the center of
the 7 oxygen atoms linking the glucopyranose units. The
results are qualitatively in agreement with the ROESY data for
intramolecular interactions. The distances between the methyl
groups (carbon atoms) and the barycenter of the CD are within
6
–16 Å for the three CD-spin traps. If we consider the C–H dis-
tance within the methyl groups (1.1 Å) and the distance
between the oxygen crown defined by the linking sugar units
and the oxygen crown of the primary rim (∼5 Å), then there is
a small range of 5–11 Å (greyscale belts, ESI†) corresponding
to a situation in which the guest methyl groups are positioned
close to the primary methoxy crown. Within this region, the
methyl groups of CD-DIPPMPO spend more time closer to the
CD barycenter (ESI†) than the methyl groups of CD-BEPMPO
which interact more frequently with the CD center with
respect to the methyl groups of CD -DEPMPO (ESI†). This
6
qualitatively reproduces the trend observed by ROESY where
the cross correlations assigned to the intramolecular associ-
ation indicate a stronger affinity for the alkyl groups of the
nitrones toward the primary methoxy groups of the CD follow-
ing the order: CD-DIPPMPO ≈ CD-BEPMPO > CD -DEPMPO.
6
The nitronyl oxygen atom was chosen as a probe near the trap-
Fig. 5 2D-ROESY spectra of (a) CD-DIPMPO, (b) CD
CD-BEPMPO and proposed types of association at the bottom (nitrone
6
-DEPMPO, and (c) ping site toward the CD center (Fig. 6) and for further compari-
son with the data obtained with the superoxide spin adducts
Fig. 7). The distance between the nitronyl oxygen atom and
the CD barycenter stays within a 6–13 Å range with a lowest
limit around 7 Å (Fig. 6, except for a short period for CD
-DEPMPO, as the main cross-peaks DEPMPO). This result shows that the polar nitrone group does
P
2 3
concentration: 20 mM, 300 K, mixing time: 400 ms, solvent: D O, CH :
(
methyl group carried by the carbon bearing the phosphoryl moiety).
CD-DIPPMPO shows mainly intramolecular association for the nitrone
6
-
5
-membered ring (blue sphere) in the vicinity of the primary methoxy
6
crown (OMe, green circle). For CD
6
are for the 5-membered ring (blue sphere) toward the secondary
not go close to the CD center and points outside of the
primary rim (mean distance ≈ 9–10 Å).
2
3
methoxy crown (OMe and OMe symbolized by yellow circles), weak
intermolecular complexes are proposed while the hydrophobicity of the
butoxy moiety of CD-BEPMPO appears to drive the association toward
the self-inclusion of the phosphorylated-nitrone part (blue drop) near
MD of superoxide spin adducts
6
the primary methoxy crown (OMe, green circle).
MD calculations were performed on the hydroperoxyl spin
adducts to estimate the position of the nitroxide function and
of the hydroperoxyl group relative to the CD (Fig. 7).
The modelling showed the structure of the CD macrocycle
to be very flexible. However, no such sugar tilting as seen for
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 6 Time evolution of the distances between the nitronyl oxygen atom guest alkyl groups and the center of permethylated β-CD (see text) for (a)
CD-DIPPMPO, (b) CD -DEPMPO and (c) CD-BEPMPO. Schemes of the most significant conformers are shown at the bottom. For CD-DIPPMPO,
around 15% of all conformers show the nitrone guest part outside the CD and around 85% have the guest moiety capping the cavity as pictured in
a). For CD -DEPMPO (b), the nitrone moiety stays mainly outside (far from the influence of the primary rim, in line with ROESY). CD-BEPMPO (c) has
6
(
6
its nitrone moiety sharing its position between an outer complex and a loose complex capping the cavity. Note the included glucopyranose
methoxy group (red arrows).
Fig. 7 (a) Time evolution of the distances between the nitroxide oxygen atom and the center of permethylated β-CD for (see text) (a)
CD-DIPPMPO-OOH, (b) CD
6
-DEPMPO-OOH and (c) CD-BEPMPO-OOH. Schemes of the most significant conformers are shown at the bottom
except for CD -DEPMPO-OOH showing that, for the time when the appended guest is near the CD, some alkyl chains can be accommodated near
6
the primary rim of the CD and the nitroxide moiety stays outside of the cavities.
This journal is © The Royal Society of Chemistry 2017
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
the corresponding nitrones was observed, which should allow
a significantly better guest accommodation near the entrance
Acknowledgements
of the cavity. Still, one (sometime two) glucopyranose unit is This work was supported by PACA Région (Apex, Fluo2) and
slightly rotated exposing its O(6)Me group (or the arm portion) ANR (ANR-09-BLAN-0193). The authors are grateful to the
near the entrance of the primary rim (Fig. 7, bottom). national EPR facilities (IR RPE CNRS 3443) and the Hungarian
However, the superoxide adduct of CD-DIPPMPO showed National Research, Development and Innovation Office
structural features similar to those of the corresponding (NKFTH) Grant (K119442).
nitrone. The methyl groups often interact with the primary rim
of the CD even though the five-membered ring (trap) stays
more often outside the CD (ESI†). This is confirmed by the
Onitroxide–CDcenter distance, which stays around 10 ± 2 Å
Notes and references
(
Fig. 7a–c top).
1 E. G. Janzen and D. L. Haire, Adv. Free Radical Chem., 1990,
1, 253.
For CD -DEPMPO, Me121 is more favorably included in the
6
CD but the Onitroxide–CDcenter distance remains large, reflecting
the positions of the nitroxide and hydroperoxyl moieties
oriented towards the bulk solvent.
2 M. J. Turner and G. M. Rosen, J. Med. Chem., 1986, 29, 2439.
3 M. J. Davies, B. C. Gilbert, C. Hazlewood and N. P. Polack,
J. Chem. Soc., Perkin Trans. 2, 1995, 13.
Similar features are observed for CD-BEPMPO (Fig. 6c and
c), the nitroxide and hydroperoxyl moieties staying outside of
4 K. Ranguelova and R. P. Mason, Magn. Reson. Chem., 2011,
49, 152.
7
the CD (Fig. 7c, bottom). Even though the nitroxides are co-
valently grafted on the CD rings, the high degrees of freedom,
especially of the CD ring, prevent efficient self-inclusion due to
the lack of cavity preorganization as found for natural β-CD or
5 M. J. Davies, Methods, 2016, 109, 21.
6 N. Beziere, M. Hardy, F. Poulhes, H. Karoui, P. Tordo,
O. Ouari, Y. M. Frapart, A. Rockenbauer, J. L. Boucher,
D. Mansuy and F. Peyrot, Free Radicals Biol. Med., 2014, 67,
150.
2
,6-di-OMe-β-CD. The breaking of the intramolecular, circular
hydrogen bond network of the secondary rim by methyl substi-
tution at position 3 is the origin of such flexibility. However,
even though the spin adduct moiety containing the nitroxide
and the hydroperoxyl functions stays outside, in agreement
with the rapid reduction of the spin adduct with ascorbate
7 C. L. Hawkins and M. J. Davies, Biochim. Biophys. Acta,
2014, 1840, 708.
8 R. P. Mason, P. M. Hanna, M. J. Burkitt and M. B. Kadiiska,
Environ. Health Perspect., 1994, 102(Suppl 10), 33.
9 B. D’Autreaux and M. B. Toledano, Nat. Rev. Mol. Cell Biol.,
2007, 8, 813.
(EPR results), the nesting of the spin trap provided by the
bulkiness of the nearby primary methoxy crown could explain 10 O. Ouari, F. Chalier, R. Bonaly, B. Pucci and P. Tordo,
the high values of half-life time observed for the superoxide
spin adducts.
J. Chem. Soc., Perkin Trans. 2, 1998, 2299.
11 M. Rosselin, F. Choteau, K. Zeamari, K. M. Nash, A. Das,
R. Lauricella, E. Lojou, B. Tuccio, F. A. Villamena and
G. Durand, J. Org. Chem., 2014, 79, 6615.
1
2 C. Frejaville, H. Karoui, B. Tuccio, F. Lemoigne, M. Culcasi,
S. Pietri, R. Lauricella and P. Tordo, J. Med. Chem., 1995,
38, 258.
Conclusions
A series of CD-spin traps have been synthesized and their
superoxide spin trapping properties evaluated. Superoxide 13 G. Olive, A. Mercier, F. Le Moigne, A. Rockenbauer and
adducts with a lifetime greater than 45 min in a buffer solu- P. Tordo, Free Radicals Biol. Med., 2000, 28, 403.
tion using the HX/XO system have been obtained, confirming 14 S. U. Kim, Y. Liu, K. M. Nash, J. L. Zweier, A. Rockenbauer
the stabilizing role of the CD macrocycle. The detection of and F. A. Villamena, J. Am. Chem. Soc., 2010, 132, 17157.
superoxide radicals using submillimolar concentrations of 15 M. Hardy, O. Ouari, L. Charles, J. P. Finet, G. Iacazio,
spin traps has been demonstrated, a result that is not achiev-
able using other spin traps such as DMPO, DEPMPO and
V. Monnier, A. Rockenbauer and P. Tordo, J. Org. Chem.,
2005, 70, 10426.
BMPO. The protection relative to bioreduction appears to orig- 16 M. Hardy, F. Chalier, O. Ouari, J. P. Finet, A. Rockenbauer,
inate from the bulkiness/crowding of the nitroxide moiety and
not from the self-inclusion in the CD cavity. The positions of
B. Kalyanaraman and P. Tordo, Chem. Commun., 2007,
1083.
the spin trap and the spin adduct moieties have been tenta- 17 M. Hardy, F. Poulhes, E. Rizzato, A. Rockenbauer,
tively estimated using 2D NMR and molecular modelling
studies. The obtained geometries indicate a modest effect of
the structural changes on the spacer arm and on the phospho-
K. Banaszak, H. Karoui, M. Lopez, J. Zielonka, J. Vasquez-
Vivar, S. Sethumadhavan, B. Kalyanaraman, P. Tordo and
O. Ouari, Chem. Res. Toxicol., 2014, 27, 1155.
nate group with respect to CD-DEPMPO. The design of 18 K. Stolze, N. Udilova, T. Rosenau, A. Hofinger and H. Nohl,
improved CD linked spin traps would require different Biol. Chem., 2003, 384, 493.
approaches, such as (i) a new chemical linker, (ii) different 19 G. Durand, B. Poeggeler, S. Ortial, A. Polidori,
positions of grafting on the CD, such as on the large rim, and
iii) changing the CD type.
F. A. Villamena, J. Boker, R. Hardeland, M. A. Pappolla and
B. Pucci, J. Med. Chem., 2010, 53, 4849.
(
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2017

View Article Online
Organic & Biomolecular Chemistry
Paper
2
0 C. Rizzi, S. Marque, F. Belin, J. C. Bouteiller, R. Lauricella, 32 D. Bardelang, L. Charles, J. P. Finet, L. Jicsinszky,
B. Tuccio, V. Cerri and P. Tordo, J. Chem. Soc., Perkin Trans.
, 1997, 2513.
H. Karoui, S. R. A. Marque, V. Monnier, A. Rockenbauer,
R. Rosas and P. Tordo, Chem. – Eur. J., 2007, 13, 9344.
2
2
1 D. G. Mitchell, G. M. Rosen, M. Tseitlin, B. Symmes, 33 Y. B. Han, B. Tuccio, R. Lauricella and F. A. Villamena,
S. S. Eaton and G. R. Eaton, Biophys. J., 2013, 105, 338. J. Org. Chem., 2008, 73, 7108.
2 J. R. Biller, M. Tseitlin, D. G. Mitchell, Z. L. Yu, 34 D. N. Polovyanenko, S. R. A. Marque, S. Lambert,
2
L. A. Buchanan, H. Elajaili, G. M. Rosen, J. P. Y. Kao,
S. S. Eaton and G. R. Eaton, ChemPhysChem, 2015, 16, 528.
3 R. P. Mason, Redox Biol., 2016, 8, 422.
L. Jicsinszky, V. F. Plyusnin and E. G. Bagryanskaya, J. Phys.
Chem. B, 2008, 112, 13157.
35 V. Ramamoorthy, A. Ramasubbu, S. Muthusubramanian
and S. Sivasubramanian, J. Inclusion Phenom. Macrocyclic
Chem., 1999, 33, 69.
2
2
4 J. Martinie, J. Michon and A. Rassat, J. Am. Chem. Soc.,
1
975, 97, 1818.
2
2
2
5 M. Lucarini and B. P. Roberts, Chem. Commun., 1996, 36 M. Spulber and S. Schlick, J. Phys. Chem. B, 2011, 115, 12415.
577. 37 I. Snyrychova, Free Radicals Biol. Med., 2010, 48, 264.
6 Y. Kotake and E. G. Janzen, J. Am. Chem. Soc., 1989, 111, 38 N. Beziere, Y. Frapart, A. Rockenbauer, J. L. Boucher,
1
5
138.
D. Mansuy and F. Peyrot, Free Radicals Biol. Med., 2010, 49,
7 Y. Sueishi, A. Miyata, D. Yoshioka, M. Kamibayashi and
437.
Y. Kotake, J. Inclusion Phenom. Macrocyclic Chem., 2010, 66, 39 K. Abbas, M. Hardy, F. Poulhes, H. Karoui, P. Tordo,
57. O. Ouari and F. Peyrot, Free Radicals Biol. Med., 2014, 71, 281.
8 M. Spulber and S. Schlick, J. Phys. Chem. A, 2010, 114, 40 X. G. Bao, P. Tao, F. A. Villamena and C. M. Hadad, Theor.
217. Chem. Acc., 2012, 131, 1–10.
9 H. Karoui, A. Rockenbauer, S. Pietri and P. Tordo, Chem. 41 G. Durand, R. A. Prosak, Y. B. Han, S. Ortial,
3
2
2
3
3
6
Commun., 2002, 3030.
0 X. Y. Song, Y. Chen and H. L. Chen, New J. Chem., 2001, 25,
A. Rockenbauer, B. Pucci and F. A. Villamena, Chem. Res.
Toxicol., 2009, 22, 1570.
9
85.
42 F. W. Hoffmann, R. J. Ess and R. P. Usingef Jr., J. Am.
Chem. Soc., 1956, 5817.
1 M. Hardy, D. Bardelang, H. Karoui, A. Rockenbauer,
J. P. Finet, L. Jicsinszky, R. Rosas, O. Ouari and P. Tordo, 43 A. Rockenbauer and L. Korecz, Appl. Magn. Reson., 1996,
Chem. – Eur. J., 2009, 15, 11114. 10, 29.
This journal is © The Royal Society of Chemistry 2017
Org. Biomol. Chem.