Substituent effect on the rate of the hydroxyl and phenyl radical spin trapping with nitrones

Article abstract of DOI:10.1246/bcsj.75.2043

Hydroxyl and phenyl radical spin trapping rates by α-phenyl-N-t-butylnitrone (PBN, N-benzylidene-t-butylamine N-oxide) and its analogs were determined using a competitive trapping method. Hydroxyl radical was generated from hydrogen peroxide in water usin

Full text of DOI:10.1246/bcsj.75.2043

c. Jpn.
© 2002 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 75, 2043–2047 (2002) 2043
e Chemical
ciety of Ja-
n
e Chemical
ciety of Ja-
n
Substituent Effect on the Rate of the Hydroxyl and Phenyl Radical Spin
Trapping with Nitrones
Trapping Rates of PBN-Type Spin Traps
Y. Sueishi et al.
Yoshimi Sueishi,* Chiharu Yoshioka, Claudio Olea-Azar,†,# Lester A. Reinke,† and Yashige Kotake††
02
43
47
SJA8
09-2673
069
Department of Chemistry, Faculty of Science, Okayama University, 3-1-1 Tsushima Naka, Okayama 700-8530
†Department of Pharmaceutical Science, University of Oklahoma Health Sciences Center,
Oklahoma City, OK 73109, U.S.A.
††Free Radical Biology and Aging Research Program, Oklahoma Medical Research Foundation,
Oklahoma City, OK 73104, U.S.A.
02
02
(Received March 18, 2002)
Hydroxyl and phenyl radical spin trapping rates by α-phenyl-N-t-butylnitrone (PBN, N-benzylidene-t-butylamine
N-oxide) and its analogs were determined using a competitive trapping method. Hydroxyl radical was generated from
hydrogen peroxide in water using UV photolysis, in the presence of the selected spin trap plus 5,5-dimethyl-pyrroline N-
oxide (DMPO). Phenyl radical was produced with UV photolysis of tetraphenyllead, and spin trapping was performed in
benzene. Spin trapping rate constants were calculated using EPR signal intensity ratios for the DMPO spin adducts vs
PBN-type spin adducts. The rate constants were strongly dependent on the kind of substituent in the spin trap; the mag-
nitude of the substituent effect was also dependent on the kind of free radicals trapped, i.e., hydroxyl or phenyl radicals.
For example, in phenyl radical trapping, the spin trapping rate in hydroxy-substituted PBNs followed Hammett’s equa-
tion, while there is no such correlation in hydroxyl radical trapping. Hydroxy-substituted PBNs such as 2-, 3-, and 4-hy-
droxy-PBNs showed much lower apparent spin trapping rates than that of PBN. The reaction between hydroxyl radical
and phenoxyl group is a likely cause.
.1
3.1
The spin trapping technique has been proved to be a useful
tool in free radical chemistry for many years.1,2 Recently, spin
trapping has also been shown to be useful in biological sys-
tems, where attention is focused on trapping of superoxide and
hydroxyl radicals.3,4 The hydroxyl radical is considered to be a
most damaging reactive oxygen species in biological systems,
and it is detectable in some cases using electron paramagnetic
resonance (EPR) spin trapping.5 However, at present only a
few spin trapping compounds (spin traps) are in practical use
for hydroxyl radical detection, because the stability of the hy-
droxyl spin adduct is usually low.6 The spin trap 5,5-dimethyl-
pyrroline N-oxide (DMPO) has been most often used for hy-
droxyl radical detection due to the unique EPR spectrum and
the stability of the spin adduct. α-Phenyl-N-t-butylnitrone
(PBN, N-benzylidene-t-butylamine N-oxide) and its substitut-
ed-phenyl analogs are not as advantageous as DMPO for the
use as hydroxyl radical spin traps, however, these compounds
have shown a wide variety of therapeutic activities in animal
disease models.7 Such activities were at least in part attributed
to the spin traps’ hydroxyl radical trapping capabilities.
which would aid in correlating the trap’s biological activity to
trapping rate. We determined hydroxyl and phenyl radical spin
trapping rates for several substituted PBN-type spin traps, us-
ing a competitive trapping method. We utilized a photolytic
method to generate free radicals, which should exert minimum
influence on the rate constants to be determined. The effects of
various substituents in PBN analogs on the hydroxyl radical
spin trapping rates are compared with those of phenyl radical,
and their mechanistic implications are discussed.
Experimental
Materials. Spin traps illustrated in Scheme 1 were used in
the present study. PBN, 4-POBN, 2-SO3-PBN, and DMPO were
purchased from Aldrich Chemical Company, Inc. Other PBN-
type spin traps and 2-Ph-DMPO were the gift from Dr. Edward G.
Janzen.8 Hydrogen peroxide (30%) and tetraphenyllead were ob-
tained from Wako Pure Chemicals, and were used as sources of
hydroxyl and phenyl radicals, respectively. Benzene and water
were purified by distillation and used as solvents.
Free Radical Generation and EPR Measurements. Hydro-
xyl and phenyl radicals were generated with UV irradiation (2 sec-
ond irradiation with a 75 W medium-pressure Mercury arc). For
instance, hydrogen peroxide (0.2 mol dm−3) and spin traps (5 ×
10−3 mol dm−3) were mixed in phosphate buffer (pH = 6.4) and
loaded in the EPR flat cell.4 After UV light was irradiated to the
sample cell outside the cavity, EPR signals were immediately re-
corded with a JEOL JES-FE3XG EPR spectrometer. Benzene so-
lution of tetraphenyllead was bubbled with nitrogen gas to deoxy-
It is speculated that in biological systems nitrones that have
higher hydroxyl radical trapping rates may show more potent
pharmacologic activity. However, this hypothesis has not been
proven yet. The objective of this study was to collect basic
data for free radical trapping rate of various nitrone spin traps,
# Present address: Department of Pharmacology, The University
of Chile, Santiago Chile

2044 Bull. Chem. Soc. Jpn., 75, No. 9 (2002)
Trapping Rates of PBN-Type Spin Traps
hydroxyl radical in the presence of two spin trapping com-
pounds (for example, PBN and DMPO) is:
hν
H2O2 → 2·OH
k1
·
DMPO + OH → ·DMPO - OH
k2
PBN + OH → ·PBN - OH
·
·
Then, the ratio of the first order rates of formation for PBN-
OH and ·DMPO-OH adducts can be expressed as follows:
d[·PBN - OH]/dt
d[·DMPO - OH]/dt
k2 [PBN]0
k1 [DMPO]0
=
,
(1)
where [PBN]0 and [DMPO]0 denote the initial concentrations
Scheme 1.
of spin traps. Thus, k2/k1 can be calculated from the concentra-
·
·
tion ratio of DMPO-OH and PBN-OH, i.e., the plot of the
concentration ratio of spin adducts against initial concentra-
tions of spin traps ([PBN]0/[DMPO]0) gives a line with the
slope k2/k1. A typical plot for the PBN/DMPO system is
shown in Fig. 2 with a straight line passing through the origin
(correlation coefficient R = 0.99), suggesting that the calcula-
tion of relative spin trapping rate constants (k2/k1) using Eq. 1
is justifiable.
genate before UV irradiation. Spin adduct concentration was cal-
culated using double integration of first-derivative EPR signal
with the aid of a computer program (WIN-RAD system, Radical
Research Inc.). The g-values were calculated using a frequency
counter (Advantest TR5214) and 2,2-diphenyl-1-picrylhydrazyl
(DPPH) as a standard.
Results and Discussion
We determined ratios of spin-trapping rate constant (k2/k1)
of hydroxyl radical for 10 PBN-type spin traps (Table 1). Pre-
viously, DMPO spin trapping rate of hydroxyl radical was de-
termined using a time-resolved spectrophotometric technique,
and was reported to be k1 = 3.6 × 109 dm3 mol−1 s−1.10 There-
fore, this value was utilized to calculate hydroxyl radical trap-
ping rate constants (k2) for the present spin traps (Table 1).
The inspection of Table 1 indicated that 2-Ph-DMPO exhib-
its a low rate in trapping hydroxyl radical as compared with
that of DMPO. This may be attributed to steric hindrance
caused by the phenyl group at 2-position, at which free radical
trapping occurs. It should be noted that ·OH trapping rate con-
stants by hydroxy-substituted PBNs, such as 2-, 3-, and 4-HO-
PBNs and SALBN, are much lower than others.
Spin Trapping of Hydroxyl Radical. Figure 1(1) shows
the typical EPR spectrum obtained in an irradiated solution of
hydrogen peroxide plus DMPO plus PBN. Two spin adduct
species are visible in this spectrum: AN = 1.53 mT and AH =
1.53 mT assigned to ·DMPO-OH (·OH spin adduct of
DMPO); and AN = 1.58 mT and AH = 0.281 mT assigned to
·PBN-OH (·OH spin adduct of PBN).3,4 Because the lowest
field lines in the EPR spectra of DMPO-OH and PBN-OH
(arrows in Fig. 1(1)) were well separated, the integrated inten-
sity of each line was used to determine the concentration ratio.
Spin trapping rates were determined using a competitive
trapping method.9 The reaction scheme for spin trapping of
·
·
Although systematic tests to determine the correlation be-
tween the spin trapping rate and pharmacologic activity have
Fig. 1. (1) EPR spectrum obtained after UV-photolysis in the
aqueous solution of hydrogen peroxide, DMPO, and PBN:
·
·
(ꢀ) DMPO-OH (g = 2.0059) and (ꢁ) PBN-OH (g =
2.0057). (2) EPR spectrum obtained after UV-photolysis
in the benzene solution of tetraphenyllead, DMPO, and
Fig. 2. The ratio of hydroxyl radical spin trapping rates for
PBN and DMPO plotted as a function of the ratio of initial
concentration of spin traps ([PBN]0/[DMPO]0): correlation
coefficient R = 0.99.
·
·
PBN: (ꢀ) DMPO-Ph (g = 2.0045) and (ꢁ) PBN-Ph (g
= 2.0054). Peaks marked with arrows were used to deter-
mine the concentration of spin adducts.

Y. Sueishi et al.
Bull. Chem. Soc. Jpn., 75, No. 9 (2002) 2045
Table 1. Hyperfine Coupling Constants (hfcc) and Rate Constants for Hydroxyl Radical Spin
Trapping at 298 K
Spin Trap
hfsc/mT
AN AH
1.51 0.178 1.48 0.02
1.62 0.536 1.34 0.02
1.53 1.53
1.56 0.230 0.85 0.01
1.58 0.281 0.71 0.01
k2/k1
10−9 k2/dm3 mol−1 s−1
log P
−0.8
4-POBN
2-SO3-PBN
DMPO
4-NO2-PBN
PBN
2-Ph-DMPO
4-CH3O-PBN 1.58 0.296 0.31 0.01
3-HO-PBN
4-HO-PBN
2-HO-PBN
SALBN
5.33
4.82
< −1
1
(k1 = 3.6 × 109 dm3 mol−1 s−1
)
−1.0
1.4
1.2
1.0
1.2
1.0
0.6
1.3
3.06
2.56
1.58
1.12
1.04
0.90
0.22
0.14
1.52
0.44 0.01
1.56 0.288 0.29 0.01
1.60 0.293 0.25 0.01
1.58 0.331 0.06 0.01
1.56 0.277 0.04 0.01
< −1
not been performed, Hamburger and McCay compared the po-
tencies of PBN, DMPO, and 4-POBN in the protection of rats
from the lethality caused by endotoxin (lipopolysccharide) ad-
ministration.11 At optimized doses, PBN showed the highest
protective action (approximately 80% protection), followed by
4-POBN (40% protection), and DMPO (25% protection), i.e.,
PBN > 4-POBN > DMPO. The apparent trapping rates for
hydroxyl radical were: 4-POBN > DMPO > PBN (Table 1),
which does not agree with the results of the animal model ex-
periments. Although a very limited comparison is possible at
present, we speculate that hydroxyl radical trapping capability
is not necessarily responsible for the pharmacologic activity in
this specific disease model. Moreover, hydrophobicity of the
spin trap may greatly modulate its tissue concentration in vivo.
The log P (where P is the 1-octanol/water partition coefficient)
values are widely used as a measure of hydrophobicity.12 In
the present system, however, the trapping rates (k2) show no
appreciable correlation to log P values (Table 1).
Since the efficiency for hydroxyl radical spin trapping of 2-
SO3-PBN is high, the very low efficiency for trapping of 2-
HO-PBN can not be accounted for by the steric hindrance due
to 2-substituent group. As can be seen in Table 1, 3-HO- and
4-HO-PBNs as well as 2-HO-PBN show low trapping efficien-
cy, in addition, the spin trapping rates in hydroxy-substituted
PBNs did not follow Hammett’s equation. Therefore, Fig. 3(1)
shows a Hammett type relationship for ·OH spin trapping of 4
monosubstituted PBN-type spin traps by excluding HO- and 2-
substituted PBNs. The reaction constant ρ for hydroxyl radical
spin trapping was 0.45.
Fig. 3. Plots of logarithms of the relative rates (kR-PBN/kPBN
)
There is a tendency that an electron-withdrawing group
such as nitro group tends to increase ·OH trapping rate, and an
electron-donating group such as methoxy group tends to de-
crease it. Previously, in a spin trapping study of t-butoxy radi-
cal, Janzen and Evans have shown the same trend in the sub-
stituent effect.2 Hirota et al.13 studied the substituted PBN spin
trapping reaction using molecular-orbital and molecular-me-
chanics calculations. They suggested that the electron densi-
ties in the lowest unoccupied molecular orbital (LUMO) of the
spin trap and the highest occupied molecular orbital (HOMO)
of the hydroxyl radical could be determinants for the trapping
reaction rate. The positive reaction constant (ρ = 0.45) for hy-
droxyl radical spin trapping reaction obtained in this study
vs Hammett σ-constants on monosubstituted PBNs (R-
PBN): (1) hydroxyl radical spin trapping and (2) phenyl
radical spin trapping.
supports the notion that the nucleophilic attack of hydroxyl
radical occurs at the double bond in the nitrone group.
Spin Trapping of Phenyl Radical. Phenyl radical spin
trapping rates by monosubstituted PBNs were determined us-
ing a competitive trapping method. Figure 1(2) shows the EPR
spectrum obtained for phenyl radical spin trapping. Hfsc’s of
the two spin adducts are: AN = 1.39 mT and AH = 1.93 mT as-
·
signed to DMPO-Ph (phenyl radical adduct of DMPO); and

2046 Bull. Chem. Soc. Jpn., 75, No. 9 (2002)
Trapping Rates of PBN-Type Spin Traps
Table 2. Hyperfine Coupling Constants (hfcc) and Rate Constants for Phenyl Radical Spin
Trapping at 298 K
Spin Trap
hfsc/mT
k2/k1
10−7 k2/dm3 mol−1 s−1
AN
AH
4-NO2-PBN
PBN
3-HO-PBN
DMPO
1.41
1.43
1.44
1.39
1.54
0.210
0.220
0.231
1.93
1.75 0.09
1.56 0.01
1.13 0.04
1
1.35
1.2
0.87
0.77
0.46
2-HO-PBN
0.179
0.60 0.03
·
·
AN = 1.43 mT and AH = 0.220 mT assigned to ·PBN-Ph (phen-
yl radical adduct of PBN).14,15 The relative spin trapping rate
constants of phenyl radical for 4 monosubstituted PBN spin
traps are listed in Table 2. Because 2-SO3- and 4-HO-PBNs
are insoluble in benzene, the rates were not measured. The ab-
solute trapping rate constant of phenyl radical by PBN was re-
ported as 1.2 × 107 dm3 mol−1 s−1 in methanol,16 and this val-
ue was used to calculate the rate constants (k2) for PBN-type
spin traps (Table 2).
where O-PBN and O-PBN(·)-OH denote phenoxyl radicals
formed by hydrogen abstraction from spin trap (HO-PBN) and
spin adduct (HO-PBN(·)-OH), respectively. The reactions
shown above consume originally added spin trap as well as
·OH radicals, and could lead to diminishing apparent trapping
efficiency by hydroxy-substituted PBN.
In a rat model, Reinke et al.18 investigated the metabolic fate
of PBN in vivo and showed that the phenyl group in PBN is
hydroxylated to form 2-, 3-, and 4-hydroxy-PBNs. A majority
(80%) of hydroxyl radical attack occurs on the aromatic ring of
PBN rather than on the nitrone group trapping. This result was
surprising, because it was thought that the hydroxyl radical at-
tack occurs primarily at the double bond in the nitrone group.
Nevertheless, hydroxy-substituted PBN should still retain free
radical-trapping capabilities and phenolic group should have
scavenging capability against hydroxyl radical. Since hydrox-
yl radicals show high reactivity with spin traps at the sites oth-
er than nitrone group sites, it is possible to speculate that free
radical trapping capability may not be a unique determinant for
the pharmacologic activity.
In conclusion, we determined spin trapping rates in various
substituted PBNs for hydroxyl radical (in water) and phenyl
radical (in benzene) using a competitive trapping method with
DMPO. We show a reason why hydroxy-substituted PBN ex-
hibits apparent low spin trapping rates. Although no correla-
tion was found between the pharmacologic activities and trap-
ping rate constants, this study may provide helpful models for
the interpretation of biological data in the future.
It is noted that, in phenyl radical spin trapping, the low trap-
ping efficiency that was seen in hydroxyl radical spin trapping
by hydroxylated PBNs was absent. Hammett-type plot for
phenyl radical trapping in benzene is shown in Fig. 3(2). The
Hammett σ-constant for 2-HO-group is not available, and thus
the σ-constant at 4-position is tentatively used. By this reason,
we excluded 2-HO-PBN from the calculation of the reaction
constant ρ, and so the slope of the Hammett plot is slightly
positive, i.e., ρ = 0.14. It is noted that the HO-trapping rates
are about 100 times faster than the Ph-trapping rates. In a ki-
netic study of superoxide and hydroxyl radicals, Finkelstein et
al.3 estimated the rate ratio for DMPO hydroxyl radical spin
·
trapping (kDMPO) and OH radical hydrogen abstraction from
ethanol (kEtOH) as kDMPO/kEtOH = 1.91. Using the kDMPO value,10
we calculate the kEtOH value as 1.89 × 109 dm3 mol−1 s−1,
·
which is comparable to the OH radical spin trapping rate of
PBN (kPBN) shown in Table 1. The ratio is calculated to be
kPBN/kEtOH = 1.35.
·OH Radical Reaction with Hydroxylated PBN. There
are many reports on hydrogen abstraction by free radicals from
OH group in phenol compounds. Hogg et al.17 studied the hy-
drogen abstraction by DPPH from phenols, and have shown
that ·OH radicals cause 1) hydrogen abstraction reaction from
the OH group in hydroxy-substituted PBNs, and 2) ·OH radi-
cal addition to the traps. Thus, after UV irradiation, the
possible ·OH radical reactions may be shown as fol-
lows:
We thank Dr. Edward G. Janzen (Belwood, Ontario Canada)
for the gift of spin traps used in this study. Claudio Olea-Azar
is grateful to the research fellowship provided by the American
Chemical Society to visit the University of Oklahoma Health
Sciences Center.
References
k1
HO· + DMPO → ·DMPO - OH
1
a) E. G. Janzen, Acc. Chem. Res., 4, 31 (1971). b) T. Doba,
k2
HO· + HO - PBN → HO - PBN(·)- OH
T. Ichikawa, and H. Yoshida, Bull. Chem. Soc. Jpn., 50, 3158
(1977). c) N. Nishimura, T. Nakamura, Y. Sueishi, and S.
Yamamoto, Bull. Chem. Soc. Jpn., 67, 165 (1994). d) Y. Sueishi
and Y. Miyake, Bull. Chem. Soc. Jpn., 70, 397 (1997).
k3
HO· + HO - PBN → ·O - PBN
k4
·O - PBN + HO - PBN → HO - PBN(·)- O - PBN
fast
2
(1973).
3
E. G. Janzen and C. A. Evans, J. Am. Chem. Soc., 95, 8205
→ Decomposition
k5
·O - PBN + HO - PBN(·)- OH → ·O - PBN(·)- OH
E. Finkelstein, G. M. Rosen, and E. J. Rauckman, J. Am.
Chem. Soc., 102, 4994 (1980).
Y. Kotake and E. G. Janzen, J. Am. Chem. Soc., 113, 9503
fast
→ Decomposition
k6
·
·O - PBN + HO - PBN → HO - PBN + O - PBN,
4

Y. Sueishi et al.
Bull. Chem. Soc. Jpn., 75, No. 9 (2002) 2047
(1991).
13 a) Y. Abe, S. Seno, K. Sakakibara, and M. Hirota, J. Chem.
Soc., Perkin Trans. 2, 1991, 897. b) K. Murofushi, K. Abe, and M.
Hirota, J. Chem. Soc., Perkin Trans. 2, 1987, 1829.
14 E. G. Janzen and J. I. Liu, J. Mag. Resonance, 9, 510
(1973).
5
6
E. G. Janzen, Free Radicals Biol., 4, 115 (1980).
E. G. Janzen and D. L. Haire, “Two decades in spin trap-
ping,” in “Advances in Free Radical Chemistry,” ed by D. D.
Tanner, JAI Press, Greenwich CN (1990), pp. 253–295.
7
Y. Kotake, Antioxid. Redox Signal, 1, 481 (1999).
R. D. Hinton and E. G. Janzen, J. Org. Chem., 57, 2646
15 E. G. Janzen and B. J. Blackburn, J. Am. Chem. Soc., 91,
4481 (1969).
16 E. G. Janzen and C. A. Evans, J. Am. Chem. Soc., 97, 205
(1975).
8
(1992).
9
G. R. Buettner and R. P. Mason, Methods Enzymol., 186,
127 (1990).
17 a) J. S. Hogg, D. H. Lohmann, and K. E. Russell, Can. J.
Chem., 39, 1588 (1961). b) N. Nishimura, T. Moriya, Y. Okino,
K. Tanabe, K. Kawabata, and T. Wakanabe, Bull. Chem. Soc. Jpn.,
50, 1969 (1977).
18 L. A. Reinke, D. R. Moore, H. Sang, E. D. Janzen, and Y.
Kotake, Free Radical. Biol. Med., 28, 345 (2000).
10 R. Sridhar, P. C. Beaumont, and E. L. Powers, J. Radio-
anal. Nucl. Chem., 101, 227 (1986).
11 S. A. Hamburger and P. B. McCay, Circ. Shock, 29, 329
(1989).
12 E. G. Janzen, M. S. West, Y. Kotake, and C. M. DuBose, J.
Biochem. Biophys. Methods, 32, 183 (1996).