Structural influence on radical formation and sensitizing capacity of alkylic limonene hydroperoxide analogues in allergic contact dermatitis

Article abstract of DOI:10.1021/tx900433n

Hydroperoxides are known to be strong contact allergens and a common cause of contact allergy. They are easily formed by the autoxidation of, for example, fragrance terpenes, compounds that are common in perfumes, cosmetics, and household products. A requirement of the immunological mechanisms of contact allergy is the formation of an immunogenic hapten-protein complex. For hydroperoxides, a radical mechanism is postulated for this formation. In our previous investigations of allylic limonene hydroperoxides, we found that the formation of carbon-and oxygen-centered radicals, as well as the sensitizing capacity, is influenced by the structure of the hydroperoxides. The aim of the present work was to further investigate the connection between structure, radical formation, and sensitizing capacity by studying alkylic analogues of the previously investigated allylic limonene hydroperoxides. The radical formation was studied in radical-trapping experiments employing 5,10,15,20-tetraphenyl- 21H,23H-porphine iron(III) chloride as an initiator and 1,1,3,3- tetramethylisoindolin-2-yloxyl as a radical trapper. We found that the investigated hydroperoxides initially form carbon-and oxygen-centered radicals that subsequently form alcohols and ketones. Trapped carbon-centered radicals and nonradical products were isolated and identified. Small changes in structure, like the omission of the endocyclic double bond or the addition of a methyl group, resulted in large differences in radical formation. The results indicate that alkoxyl radicals seem to be more important than carbon-centered radicals in the immunogenic complex formation. The sensitizing capacities were studied in the murine local lymph node assay (LLNA), and all hydroperoxides tested were found to be potent sensitizers. For two of the hydroperoxides investigated, the recently suggested thiol-ene reaction is a possible mechanism for the formation of immunogenic complexes. For the third investigated, fully saturated, hydroperoxide, the thiol-ene mechanism is not possible for immunogenic complex formation. This strongly indicates that several radical reaction pathways for immunogenic complex formation of limonene hydroperoxides are active in parallel.

Full text of DOI:10.1021/tx900433n

Chem. Res. Toxicol. 2010, 23, 677–688
677
Structural Influence on Radical Formation and Sensitizing Capacity
of Alkylic Limonene Hydroperoxide Analogues in Allergic Contact
Dermatitis
Staffan G. H. Johansson,† Katarina Emilsson,† Morten Grøtli,‡ and Anna Bo¨rje*,†
Dermatochemistry and Skin Allergy and Medicinal Chemistry, Department of Chemistry, UniVersity of
Gothenburg, SE-412 96 Gothenburg, Sweden
ReceiVed December 4, 2009
Hydroperoxides are known to be strong contact allergens and a common cause of contact allergy.
They are easily formed by the autoxidation of, for example, fragrance terpenes, compounds that are
common in perfumes, cosmetics, and household products. A requirement of the immunological mechanisms
of contact allergy is the formation of an immunogenic hapten-protein complex. For hydroperoxides, a
radical mechanism is postulated for this formation. In our previous investigations of allylic limonene
hydroperoxides, we found that the formation of carbon- and oxygen-centered radicals, as well as the
sensitizing capacity, is influenced by the structure of the hydroperoxides. The aim of the present work
was to further investigate the connection between structure, radical formation, and sensitizing capacity
by studying alkylic analogues of the previously investigated allylic limonene hydroperoxides. The radical
formation was studied in radical-trapping experiments employing 5,10,15,20-tetraphenyl-21H,23H-porphine
iron(III) chloride as an initiator and 1,1,3,3-tetramethylisoindolin-2-yloxyl as a radical trapper. We found
that the investigated hydroperoxides initially form carbon- and oxygen-centered radicals that subsequently
form alcohols and ketones. Trapped carbon-centered radicals and nonradical products were isolated and
identified. Small changes in structure, like the omission of the endocyclic double bond or the addition of
a methyl group, resulted in large differences in radical formation. The results indicate that alkoxyl radicals
seem to be more important than carbon-centered radicals in the immunogenic complex formation. The
sensitizing capacities were studied in the murine local lymph node assay (LLNA), and all hydroperoxides
tested were found to be potent sensitizers. For two of the hydroperoxides investigated, the recently
suggested thiol-ene reaction is a possible mechanism for the formation of immunogenic complexes. For
the third investigated, fully saturated, hydroperoxide, the thiol-ene mechanism is not possible for
immunogenic complex formation. This strongly indicates that several radical reaction pathways
for immunogenic complex formation of limonene hydroperoxides are active in parallel.
electrophiles that participate in nonradical reactions. For urushi-
ols (4, 5) and hydroperoxides (6-9), it has been proposed that
the formation of immunogenic complexes can take place through
a radical mechanism. In the case of urushiols, this is proposed
as an alternative to a mechanism proceeding via the oxidation
to quinones (5). The formation of radicals from hydroperoxides
has been shown to cause unspecific oxidation of proteins (10).
In theory, this unspecific oxidation could generate the im-
munogenic complexes of hydroperoxides. If so, hydroperoxides
with different structures would form the same immunogenic
complexes, and individuals sensitized to one hydroperoxide
would react to other hydroperoxides (cross-reactivity). However,
it has been shown that neither 1-(1-hydroperoxyl-1-methylethyl)
cyclohexene (cyclohexene-hydroperoxide) (1, Figure 1) nor
cumene hydroperoxide (2) cross-react with limonene-2-hydro-
peroxide (3) in guinea pigs, although they cross-react with each
other (11). Taken together, these results show that the structure
of the hydroperoxide is important for the formation of specific
immunogenic complexes of hydroperoxides, and although
unspecific oxidation of proteins does occur in the presence of
radicals derived from hydroperoxides (10), there is no evidence
for the formation of specific immunogenic complexes via this
mechanism. Furthermore, the combination of radical-trapping
experiments (12), cross-reactivity studies, and computational
calculations (11) indicates that the active hapten of 1 is a mixture
Introduction
Allergic contact dermatitis (ACD)1 is the clinical manifesta-
tion of contact allergy. It is estimated that 15-20% of the
population in the Western world are allergic to one or more
chemicals in their environment (1). A prerequisite for the
immunologic mechanisms of contact allergy is that the com-
pounds causing the contact allergy react with macromolecules
in the skin (e.g., proteins). The allergenic compounds, usually
small organic molecules or metal ions, are known as haptens.
The reaction results in the formation of an immunogenic
hapten-protein complex. This complex is specific to each
hapten and recognized by the immune system, thereby starting
an immunological process that ultimately leads to the develop-
ment of contact allergy and ACD (2).
The most common mechanism for the immunogenic complex
formation is the nucleophilic attack by an amino acid side chain
on an electrophilic hapten (3). This mechanism is valid for
* To whom correspondence should be addressed. Tel: +46 31 786 9012.
Fax: +46 31 772 3840. E-mail: aborje@chem.gu.se.
† Dermatochemistry and Skin Allergy.
‡ Medicinal Chemistry.
1 Abbreviations: ACD, allergic contact dermatitis; Fe(III)TPPCl, 5,10,15,20-
tetraphenyl-21H,23H-porphine iron(III) chloride; LLNA, local lymph node
assay; EC3, estimated concentration required to produce a stimulation index
of 3; TMIO, 1,1,3,3-tetramethylisoindolin-2-yloxyl.
10.1021/tx900433n  2010 American Chemical Society
Published on Web 02/17/2010

678 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Johansson et al.
Scheme 1. Radical Reaction Pathways of Allylic (Right) and
Alkylic (Left) Oxygen-Centered Radicalsa
a n.r., no reaction.
Scheme 2. Synthetic Routes to Hydroperoxides 12, 14, and 17
Figure 1. Structures of compounds referred to in this paper. Compounds
1 and 2 were used in a previous study (11). R-Limonene is a commonly
used fragrance terpene that forms hydroperoxides 3 and 4 on air
exposure (17). The sensitizing capacities and radical formation of allylic
hydroperoxides 3, 4, and 5 have been previously investigated (20). The
present study investigates alkylic hydroperoxides 12, 14, and 17, using
TMIO as a radical trapper. Attempted synthesis of hydroperoxide 23
was unsuccessful.
of carbon-centered and oxygen-centered radicals, whereas it is
mainly an oxygen-centered alkoxyl radical in the case of 2. This
suggests the importance of both carbon- and oxygen-centered
radicals in the formation of immunogenic hapten-protein
complexes of hydroperoxides.
Terpenes are frequently used in household and cosmetics
products due to their pleasant scent. In the presence of air and
an initiator such as metal ions, heat, or ultraviolet light, these
compounds are autoxidized to form a range of oxidation
products (13-16). The autoxidation is facilitated by the presence
of allylic hydrogens, and the primary oxidation products are
allylic hydroperoxides; these have repeatedly been shown to
be strong contact allergens (15-18).
R-Limonene, one of the most commonly used fragrance
terpenes (19), is autoxidized to form hydroperoxides 3 and 4
(Figure 1). For these hydroperoxides and a synthetic analogue
(5), we have previously shown that both carbon- and oxygen-
centered radicals are available for the formation of immunogenic
complexes (20). Alkoxyl radicals corresponding to each hydro-
peroxide are formed when the oxygen-oxygen bond is cleaved
homolytically. These alkoxyl radicals react according to three
major pathways: hydrogen abstraction, 1,2-shift, and 1,3-
cyclization (Scheme 1). The preference for the different
pathways is governed by the structure of the hydroperoxide.
However, all of the hydroperoxides formed carbon-centered
radicals through 1,3-cyclization of the initially formed alkoxyl
radical involving the endocyclic double bond (20).
previous study, whereas the hydrogen abstraction and the 1,2-shift
are still possible reaction pathways (Scheme 1). A possible reaction
not observed in the previous study is ꢀ-scission, which could
generate cyclic ketones as well as open or expand the cyclohexane
ring. The opening or expansion of the cyclohexane ring would
generate carbon-centered radicals that would be available for the
formation of immunogenic hapten-protein complexes.
The aim of the present study was to investigate how radical
formation and sensitizing capacity was influenced by the omission
of the endocyclic double bond of the three previously studied allylic
limonene hydroperoxides 3-5 (Figure 1). Removal of the endocy-
clic double bond will prevent the 1,3-cyclization observed in the
To perform said investigations, alkylic analogues to the allylic
hydroperoxides 3-5 were synthesized (Scheme 2) and subjected
to radical-trapping studies. These were performed in the same
manner as in our previous study of hydroperoxides 3-5 (20),
employing Fe(III)TPPCl as a radical initiator and TMIO as a

Structural Influence on Sensitizing Capacity
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 679
water (solvent A) and 0.1% formic acid in acetonitrile (solvent B),
employing an isocratic system with 50% solvent A in B. The API-
ES interface was used in the positive mode, and the spray chamber
settings were as follows: nebulizer pressure, 40 psig; capillary
voltage, 4500 V; drying gas temperature, 150 °C; and drying gas
flow rate, 12 L/min. The fragmentor voltage was set to 50 V, and
the mass spectrometer was used in the scan mode.
radical trapper (Scheme 3). Trapped carbon-centered radicals
and nonradical products were isolated and identified using NMR
and MS. The sensitizing capacity was investigated in the murine
local lymph node assay (LLNA).
Experimental Procedures
GC analyses were performed on a Hewlett-Packard 6890 gas
chromatograph equipped with an on-column injector and a flame
ionization detector, using a 30 m fused silica column (HP-5; i.d.
0.25 mm, 0.25 µm film thickness) and nitrogen as the carrier gas.
The column temperature was 35 °C at injection, held isothermally
for 2 min, raised to 185 °C at a rate of 5 °C/min, and finally held
at 185 °C for 5 min. The detector temperature was 250 °C, and
1,2,3,5-tetramethylbenzene was used as the internal standard.
Column chromatography was performed using Merck silica gel
60 (230-400 mesh ASTM), and TLC was performed using silica
plated aluminum sheets (Merck, 60 F254 silica gel) that were
developed with an anisaldehyde dip (2.1 mL of acetic acid, 5.1
mL of anisaldehyde, and 7 mL of H2SO4 in 186 mL of ethanol)
followed by heating.
Melting points were recorded on a Bu¨chi Melting Point B-545
and are uncorrected. Elemental analysis (EA) was performed by
H. Kolbe Mikroanalytisches Laboratorium (Mu¨lheim an der Ruhr,
Germany). High-resolution mass spectra (HRMS) were obtained
from the Department of Chemistry at Lund University (Lund,
Sweden).
Caution: This study inVolVes skin-sensitizing compounds, which
must be handled with care.
Chemicals. (2R,5R)-5-Isopropenyl-2-methylcyclohexanone [(+)-
dihydrocarvone, Fluka, purum 77%], p-toluenesulfonhydrazide
(97%), sodium borohydride (98%), hydrogen peroxide (30-35%
in water), triethylsilane (97%), sodium peroxide, 5,10,15,20-
tetraphenyl-21H,23H-porphine iron(III) chloride (Fe(III)TPPCl,
97%), cobalt(II) chloride, 2,2,6,6-tetramethyl-3,5-heptanedione
(g98%), tert-butyl hydroperoxide (5.5 M in decane), and N-
benzylphthalimide (99%) were purchased from Sigma Aldrich
(Stockholm, Sweden). Sodium bis(trimethylsilyl)amide (2 M in
THF) and n-butyllithium (2.5 M in hexanes) were purchased from
Acros Organics (Geel, Belgium). Methyltriphenylphosphonium
bromide (98%) was purchased from Lancaster (Lancashire, United
Kingdom) and Sigma Aldrich. 4-Isopropylcyclohexanone (16) was
purchased from Frinton Laboratories (Vineland, NJ), acetone was
purchased from Merck (Darmstadt, Germany), and olive oil
was purchased from Apoteket AB (Gothenburg, Sweden).
Instrumentation and Mode of Analysis. NMR spectroscopy
was performed on a JEOL Eclipse+ 400 instrument at 400 MHz
using CDCl3 as the solvent. Chemical shifts (δ) are reported in
Synthesis. 1,1,3,3-Tetramethylisoindolin-2-yloxyl [TMIO (21)],
(4R)-4-isopropenyl-1-methyl-2-methylenecyclohexane [6 (22)], (5R)-
5-isopropenyl-1,2-dimethyl-cyclohexane-1-ol [7 (23)], and 4-iso-
propenyl-1-methyl-cyclohexanol [8 (24)] were synthesized accord-
ing to the literature.
1
1
ppm relative to CHCl3 at 7.26 for H and at 77.0 ppm for 13C. H
and 13C NMR spectra were assigned using 13C distortionless
enhancement by polarization transfer (DEPT), H-1H correlation
1
spectroscopy (COSY), 1H-13C heteronuclear multiple quantum
coherence (HMQC), and 1H-13C heteronuclear multiple bond
correlation (HMBC).
(5R)-5-Isopropenyl-2-methyl-cyclohexane-1-p-tosylhydrazone (9).
(5R)-5-Isopropenyl-2-methyl-cyclohexanone (10, 2.0 mL, 12.2
mmol) was added to a stirred solution of p-toluenesulfonhydrazide
(2.7 g, 14 mmol) in ethanol (20 mL) at 50 °C. The reaction was
followed by TLC until all of the starting material was consumed.
The reaction mixture was concentrated under reduced pressure,
dissolved in chloroform, and absorbed on silica. The silica was
washed with ethyl acetate:hexane (1:3), and the solvent was
concentrated under reduced pressure affording 9 as a yellow oil.
1H NMR: δ 0.99-1.03 (m, 3H, H10), 1.13-1.18 (m, 1H, H3 or
H4 or H6), 1.31-1.43 (m, 1H, H3 or H4 or H6), 1.62 (m, 1H, H3
or H4 or H6), 1.66 (s, 3H, H9), 1.76-1.83 (m, 1H, H3 or H4 or
H6), 1.87-1.96 (m, 1H, H3 or H4 or H6), 1.96-2.01 (m, 1H, H5),
2.06-2.15 (m, 1H, H2), 2.41 (s, 3H, Ar-CH3), 2.65-2.73 (m, 1H,
H3 or H4 or H6), 4.64-4.72 (m, 2H, H8), 7.26-7.31 (m, 2H, Ar-
H), 7.50 (br s, 1H, H16), 7.81-7.87 (m, 2H, Ar-H). 13C NMR: δ
16.4 (C10), 20.7 (C9), 21.7 (Ar-CH3), 31.0 (C3 or C4 or C6), 31.9
(C3 or C4 or C6), 35.3 (C3 or C4 or C6), 39.3 (C2), 45.2 (C5),
109.8 (C8), 128.4 [2C, SC(CH)2], 129.3 [2C, (CH)2CCH3], 135.3
(Ar-CH3), 143.9 (C11), 148.0 (C7), 163.6 (C1). MS (API-ES, 120
eV) m/z (%): 343 [M + Na] (10), 321 [M + H] (100), 166 (13),
151 (18), 107 (18). Anal. calcd for C17H24: C, 63.72; H, 7.55. Found:
C, 63.96; H, 7.74. The product was used without further purification
in the synthesis of 11.
(5R)-5-Isopropenyl-2-methyl-cyclohexane-1-p-tosylhydrazine (11).
Sodium borohydride (0.13 g, 3.4 mmol) was slowly added to a
stirred mixture of 9 (0.54 g, 1.7 mmol) in dichloromethane (28
mL) at 0 °C under N2. After 10 min, anhydrous methanol (7 mL)
was added, and the solution was stirred at 0 °C until HPLC showed
no starting material. The reaction was quenched by slow addition
of water (15 mL) followed by brine (10 mL), and the aqueous phase
was extracted with dichloromethane (5 × 20 mL). The organic
phase was washed with brine (2 × 20 mL), dried over MgSO4,
and concentrated under reduced pressure affording 11 as a yellow
oil that was used without further purification.
Analytical HPLC was performed using two Gilson pumps model
305, a BAS UV/vis detector model UV-116 (Bioanalytical Systems,
Inc., West Lafayette, IN), a CMA 200 Microsampler (CMA
Microdialys AB, Stockholm, Sweden), and a Zorbax Rx-SiL column
(250 mm × 4.6 mm i.d., particle size 5 µm, Agilent). The mobile
phase consisted of 40% tert-butyl methyl ether in hexane, the flow
rate was 1.0 mL/min, and the compounds were monitored at
254 nm.
Preparative HPLC was performed using a Gilson pump model
305, a Gilson UV/vis detector model 119 (Gilson Medical Electron-
ics, Inc., Middleton, WI), and a Zorbax Rx-SiL prepHT column
(250 mm × 21.2 mm i.d., particle size 7 µm, Agilent). Various
concentrations of tert-butyl methyl ether in hexane (specified below)
were used as the mobile phase, the flow rate was 21.24 mL/min,
and the compounds were monitored at 205 nm.
LC/MS analyses were performed on a Hewlett-Packard 1100
HPLC/MS including a vacuum degasser, a binary pump, an
autoinjector, a column thermostat, a DAD detector, and a single
quadrupole mass spectrometer. The HPLC was equipped with a
Zorbax SB-C18 column (150 mm × 3.0 mm i.d., particle size 3.5
µm, Agilent), and the mobile phase consisted of 0.005% pentafluo-
ropentanoic acid, 0.1% acetic acid, and 5% acetonitrile in water
(solvent A) together with 0.005% pentafluoropentanoic acid, 0.1%
acetic acid, and 5% water in acetonitrile (solvent B). A linear
gradient of solvent B 0-100% for 25 min was followed by 10 min
of isocratic elution with 100% B; this was followed by a 3 min
linear gradient 100-0% B and ended with a 3 min isocratic elution
with 100% solvent A. The column temperature was 40 °C, and the
flow rate was 0.40 mL/min. The mass spectrometer was equipped
with an atmospheric pressure ionization electrospray (API-ES)
interface used in the positive ionization mode with the following
settings in the spray chamber: nebulizer pressure, 40 psig; capillary
voltage, 3500 V; drying gas temperature, 350 °C; and drying gas
flow rate, 10 L/min. The fragmentor voltage was set to 50-120 V,
and the mass spectrometer was used in both the scan mode and the
selected ion monitoring (SIM) mode. Flow injection analysis (FIA)
of the hydroperoxides was performed using 0.1% formic acid in
(5R)-5-Isopropenyl-2-methyl-cyclohexane-1-hydroperoxide (12).
Because of safety precautions, two round-bottomed flasks were each
charged with 11 (3.3 g, 10 mmol) in THF (100 mL). To each flask,
aqueous H2O2 (30%, 103 mL, 1 mol) was slowly added at 0 °C

680 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Johansson et al.
followed by addition of Na2O2 (1.2 g, 15 mmol). The solutions
were slowly warmed to room temperature and stirred until TLC
showed no starting material. The reaction mixtures were pooled,
diluted with water (200 mL), and acidified with 2 M HCl to pH
5-6, whereafter brine (30 mL) was added, and the solution was
extracted with dichloromethane (3 × 200 mL). The organic phases
were pooled, washed with water (2 × 200 mL), dried over Na2SO4,
and concentrated under reduced pressure at room temperature
affording 12 as a yellow oil (mixture of diastereomers). The
diastereomers were enriched by two rounds of preparative HPLC
purifications (isocratic, hexane/tert-butyl methyl ether; first round,
85:15; second round, 95:5), yielding diastereomers 12a (0.22 g,
14%), 12b (0.12 g, 7.8%), and 12c (0.11 g, 7.3%). The fourth
diastereomer (12d) could not be enriched to be a main component
in any fraction using these systems. Compound 12a (major isomer):
1H NMR: δ 0.98 (d, 3H, J ) 6.7 Hz, H10), 1.27-1.38 (m, 2H, H3
or H4 or H6), 1.42-1.46 (m, 1H, H3 or H4 or H6), 1.46-1.52 (m,
1H, H3 or H4 or H6), 1.69 (m, 1H, H2), 1.70 (s, 3H, H9),
1.72-1.74 (m, 1H, H3 or H4 or H6), 1.84-1.93 (m, 1H, H3 or H4
or H6), 2.18-2.26 (m, 1H, H5), 3.93-3.96 (m, 1H, H1), 4.65-4.74
(m, 2H, H8), 8.01 (s, 1H, H11). 13C NMR: δ 17.0 (C10), 21.3 (C9),
25.7 (C3 or C4 or C6), 27.3 (C3 or C4 or C6), 29.5 (C3 or C4 or
C6), 30.4 (C2), 39.0 (C5), 85.6 (C1), 109.0 (C8), 149.5 (C7). MS
(API-ES, 50 eV) m/z (%): 193 [M + Na] (23), 171 [M + H] (86),
153 (100), 137 (16), 135 (36), 64 (47). Anal. calcd for C10H18O2:
C, 70.55; H, 10.66; O, 18.67. Found: C, 70.59; H, 10.62; O, 18.67.
For the characterization data of minor isomers, see the Supporting
Information.
(m, 2H, H4), 2.18-2.28 (m, 2H, H5 and HA6), 4.69 (s, 2H, H8),
6.96 (s, 1H, OOH). 13C NMR: δ 15.0 (C10), 21.2 (C9), 23.3 (C11),
30.6 (C3), 31.8 (C4), 39.6 (C6), 40.1 (C5), 40.8 (C2), 82.7 (C1),
108.4 (C8), 150.3 (C7). MS (API-ES, 50 eV) m/z (%): 207 [M +
Na] (36), 185 [M + H] (46), 167 (100), 151 (21), 149 (70), 83
(75), 64 (82). Anal. calcd for C11H20: C, 71.70; H, 10.94. Found:
C, 71.66; H, 11.03.
1-Methylene-4-isopropyl-cyclohexane (15). Butyllithium (1.0 mL,
2.5 M) was added dropwise to a stirred suspension of methyltriph-
enylphosphonium bromide (0.9 g, 2.6 mmol) in freshly distilled
THF (12 mL) under N2 at 0 °C. After 15 min, 16 (0.3 g, 2.1 mmol)
was added dropwise, and the reaction mixture was slowly warmed
to room temperature. After 5.5 h, pentane (10 mL) and water (10
mL) were added, the phases were separated, the aqueous phase
was extracted with pentane (10 mL), and the organic phases were
pooled, washed with water (2 × 25 mL), and filtered to separate a
solid byproduct. The filtrate was dried over Na2SO4, filtered, and
concentrated under reduced pressure at 0 °C yielding a colorless
oil that was used without further purification. NMR data of the
crude product correspond to the literature (22).
4-Isopropyl-1-methyl-cyclohexane-1-hydroperoxide (17). Trieth-
ylsilane (0.5 mL, 3.2 mmol) and t-butyl hydroperoxide (one drop,
5.5 M in decane) were added to a stirred solution of 15 (0.2 g, 1.6
mmol) in 1,2-dichloroethane (8.9 mL) under oxygen atmosphere
at room temperature. Co(thd)2 (22 mg, 53 µmol) was added to the
clear solution, whereupon it instantly turned dark green. The mixture
was stirred for 1 h and 40 min and filtered through silica eluting
with dichloromethane. The dichloromethane was evaporated,
methanol (1 mL) and aqueous HCl (1 M, 1 drop) were added, and
the solution was stirred at room temperature for 30 min. Dichlo-
romethane (10 mL) and water (10 mL) were added to the reaction
mixture, the phases were separated, and the aqueous phase was
extracted with dichloromethane (10 mL). The organic phases were
pooled, washed with water (25 mL), dried over Na2SO4, filtered,
and concentrated under reduced pressure. The crude product was
purified using flash chromatography on silica gel eluting with
hexane/ethyl acetate (9:1) yielding 0.87 g (23% from 16) of the
Bis(2,2,6,6-tetramethyl-3,5-heptanedionato)cobalt(II) [Co(thd)2].
Cobalt(II) chloride (2.2 g, 17 mmol) was added to a stirred solution
of 2,2,6,6-tetramethyl-3,5-heptanedione (6.3 g, 34 mmol) in deion-
ized water (30 mL) under nitrogen atmosphere at 65 °C. Sodium
hydroxide (1.4 g, 34 mmol in 10 mL of H2O) was added to the
pink solution, whereupon it turned blue, and pink crystals im-
mediately precipitated. After 3 h, the aqueous mixture was filtered,
and the solid was dissolved in diethyl ether (100 mL), filtered
through Celite, and concentrated under reduced pressure. The pink
crude product was sublimed at 90 °C yielding Co(thd)2 {6.4 g, 44%,
mp 142 °C [lit. 143 °C (25)]} as purple crystals.
1
target compound. H NMR δ 0.85 (d, J ) 6.96 Hz, 6H, H9 and
H10), 0.94-1.04 (m, 1H, H8), 1.21 (s, 3H, H7), 1.23-1.28 (m,
4H, HA2 and HA3 and HA5 and HA6), 1.39-1.49 (m, 3H, HB3 and
H4 and HB5), 1.89-2.01 (m, 2H, HB2 and HB6). 13C NMR: δ 20.0
(C9 and C10), 24.9 (C3 and C5), 25.4 (C7), 34.5 (C4), 35.5 (C2
and C6), 43.5 (C8), 80.9 (C1). MS (API-ES, 50 eV) m/z (%): 195
[M + Na] (16), 173 [M + H] (2), 155 (100), 139 (13), 83 (17), 81
(17), 74 (51), 64 (20). Anal. calcd for C10H20O2: C, 69.72; H, 11.70.
Found: C, 69.70; H, 11.88.
(5R)-5-Isopropenyl-1,2-dimethyl-cyclohexane-1-triethylsilyl per-
oxide (13). Triethylsilane (4.4 mL, 27.6 mmol) and t-butyl hydro-
peroxide (5.5 M in decane, 0.69 µmol, 125 µL) were added to a
stirred solution of 6 (2.1 g, 13.8 mmol) in 1,2-dichloroethane (125
mL) under oxygen atmosphere. Co(thd)2 (0.15 g, 0.36 mmol) was
added to the clear solution at room temperature, whereupon it
instantly turned dark green. The mixture was stirred for 20 min,
filtered through silica, and concentrated. The crude product was
purified by liquid chromatography (100% hexane) yielding 13 (0.35
4-Isopropyl-1-methyl-cyclohexanol (18). Pd/C (80 mg, 5%) was
added to a solution of 8 (0.38 g, 2.5 mmol) in ethanol (17 mL).
The mixture was stirred at room temperature under an atmosphere
of H2 for 6 h, exposed to air overnight, filtered through Celite, and
concentrated under reduced pressure. The crude product was purified
by liquid chromatography (hexane/ethyl acetate 85:15) yielding two
1
g, 8.7%) as a clear oil. H NMR: δ 0.62-0.70 [q, J ) 8.06 Hz,
6H, Si(CH2CH3)3], 0.90-0.93 (d, J ) 6.22 Hz, 3H, H10),
0.93-0.95 (m, 1H, HA6), 0.95-1.01 [t, J ) 8.06 Hz, 9H,
Si(CH2CH3)3], 1.22 (s, 3H, H11), 1.39-1.46 (m, 4H, H2, HA3, H4),
1.72 (s, 3H, H9), 1.73-1.76 (m, 1H, HB3), 2.21-2.35 (m, 2H, H5
and HB6), 4.66 (s, 2H, H8). 13C NMR: δ 4.0 [Si(CH2CH3)3], 6.9
[Si(CH2CH3)3], 15.1 (C10), 21.3 (C9), 23.5 (C11), 30.9 (C4), 31.7
(C3), 39.9 (C5), 40.1 (C6), 41.1 (C2), 82.0 (C1), 107.8 (C8), 151.0
(C7). MS (API-ES, 50 eV) m/z (%): 321 [M + Na] (2), 299 [M +
H] (11), 183 (49), 167 (46), 151 (31). Anal. calcd for C17H34O2Si:
C, 68.39; H, 11.48. Found: C, 68.12; H, 11.45.
1
diastereomers of 18 as clear oils. Major isomer (0.14 g, 37%): H
NMR: δ 0.92 (d, J ) 6.96 Hz, 6H, H9 and H10), 0.97-1.07 (m,
1H, H8), 1.25 (s, 3H, H7), 1.30-1.39 (m, 2H, HA3 and HA5),
1.35-1.46 (m, 2H, HA2 and HA6), 1.44-1.54 (m, 1H, H4),
1.54-1.62 (m, 2H, HB3 and HB5), 1.67-1.75 (m, 2H, HB2 and
HB6). 13C NMR: δ 20.1 (C9 and C10), 25.2 (C3 and C5), 31.5
(C7), 32.8 (C4), 39.1 (C2 and C6), 43.6 (C8), 69.3 (C1). MS (API-
ES, 100 eV) m/z (%) 139 (29), 124 (14), 97 (18), 83 (100), 71
(17), 69 (18), 60 (16). HRMS (ESI) m/z calcd for C10H20O,
156.1514; found, 156.1510. For the characterization data of the
minor isomer (62 mg, 16%), see the Supporting Information.
(5R)-5-Isopropenyl-1,2-dimethyl-cyclohexane-1-hydroperoxide
(14). Compound 13 (0.33 g, 11.1 mmol) was swirled in a solution
of 2 drops of concentrated HCl in methanol (10 mL) for 2 min.
The solution was washed with sodium bicarbonate (saturated, 50
mL) and extracted with diethyl ether (2 × 50 mL). The pooled
organic phases were washed with brine (50 mL) and dried over
MgSO4, filtered, and concentrated under reduced pressure at room
temperature. The crude product was purified by liquid chromatog-
raphy (hexane/ethyl acetate 9:1) yielding 14 (0.15 g, 73%) as a
clear oil. 1H NMR: δ 0.92-0.96 (d, 3H, H10, J ) 6.59 Hz),
1.04-1.13 (m, 1H, HA6), 1.19-1.22 (m, 1H, HA3), 1.29 (s, 3H,
H11), 1.45-1.47 (m, 2H, H2 and HB3), 1.73 (s, 3H, H9), 1.74-1.81
Chemical Radical-Trapping Experiments with TMIO, General
Procedure. Fe(III)TPPCl (1 equiv) was added to a solution of
hydroperoxide (1 equiv) and TMIO (2 equiv) in acetonitrile/water
(1:1). The reaction was stirred at room temperature until TLC
showed no starting material. The reaction mixture was filtered to
remove excess Fe(III)TPPCl, and the residue was washed three
times with hexane. The phases were separated, the aqueous phase
was extracted three times with hexane, and the organic phases were

Structural Influence on Sensitizing Capacity
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 681
pooled, dried over MgSO4, and concentrated under reduced pressure.
The crude products were purified using chromatography on silica
gel eluting with various mixtures of solvents specified below. The
isolated products were characterized and quantified using 1D and
2D NMR, LC/MS, GC/FID, and EA.
Trapping Experiments with 12. Starting from 0.31 g (1.8 mmol)
of 12, two products were isolated using liquid chromatography on
silica gel eluting with stepwise gradient hexane/diethyl ether
(1:1), hexane/ethyl acetate (1:1), hexane/methanol (9:1). The
products isolated were dihydrocarvone (10, 0.11 g, 41%) and
dihydrocarveol (19, 8.93 mg, 3.2%). 1H and 13C NMR data agreed
with authentic samples.
Table 1. LLNA Responses for Compounds 3-5, 12, 14, and 17a
EC3 value
compd and test
concn (% w/v)
[3H]thymidine incorp
(dpm/lymph node)
SI
M
% w/v
3b
0.049
0.019
0.071
0.065
0.83
0.33
1.29
1.1
4b
5b
12
control
0.1
0.5
1.0
2.5
7.5
14
control
0.1
0.5
1.0
2.5
7.5
17
702
902
1010
1910
4724
9908
1.3
1.4
2.7
6.7
14
Trapping Experiments with 14. Starting from 0.29 g (1.6 mmol)
of 14, three products were isolated using liquid chromatography
on silica gel eluting with stepwise gradient hexane/diethyl ether
(1:1), hexane/ethyl acetate (1:1), hexane/methanol (9:1). The
products isolated were compounds 20, 21, and 7. For 4′-isopropenyl-
7′-(1,1,3,3-tetramethylisoindolyl-2-oxy)-octa-2′-one (20, 0.26 g,
47%, mixture of diastereomers 50:50): 1H NMR: δ 1.20-1.24 (m,
3H, H8′), 1.26-1.32 (m, 4H, HA5′ and H10 or H11 or H12 or
H13), 1.32-1.38 (m, 4H, HB5′ and H10 or H11 or H12 or H13),
1.40-1.54 (m, 6H, H10 and/or H11 and/or H12 and/or H13),
1.63-1.66 (m, 2H, H6′), 1.67 (s, 3H, H11′), 2.12 (s, 3H, H1′),
2.46-2.56 (m, 2H, H3′), 2.58-2.66 (m, 1H, H4′), 3.83-3.89 (m,
1H, H7′), 4.72-4.80 (m, 2H, H10′), 7.06-7.12 (m, 2H, H7 and
H4), 7.18-7.24 (m, 2H, H5 and H6). 13C NMR: δ 18.8/18.9 (C11′),
19.9/20.0 (C8′), 25.3/25.5 (2C, C10 or C11 or C12 or C13), 29.2/
29.3 (C5′), 30.2/30.4 (C1′), 30.7 (2C, C10 or C11 or C12 or C13),
33.66/33.72 (C6′), 43.05/43.14 (C4′), 48.6 (C3′), 67.3 (C1 or C3),
67.7 (C1 or C3), 78.6/78.8 (C7′), 112.17/112.21 (C10′), 121.6 (C4
and C7), 127.2 (C5 and C6), 145.47/145.54 (C9′), 146.47/146.54
(C8 and C9), 208.3 (C2′). MS (API-ES, 120 eV) m/z (%): 380 [M
+ Na] (1), 359 (25), 358 [M + H] (100), 192 (30), 161 (6), 160
(48). Anal. calcd for C23H35NO2: C, 77.27; H, 9.87. Found: C, 77.72;
H, 9.73. For 4-isopropenyl-octa-2,7-dione (21, 40.4 mg, 14%),
0.037
0.020
0.68
0.35
509
734
1036
2378
7092
6848
1.4
2.0
4.7
14
14
control
0.1
0.5
1.0
2.5
7.5
226
217
968
1643
1812
3927
1.0
4.3
7.3
8.0
17
a The local lymph node experiments were performed as previously
reported (20). The SIs correspond to the increase in thymidine
incorporation of treated groups relative to vehicle-treated controls. EC3
values (the estimated concentration required to induce an SI of 3) were
calculated using linear interpolation (40). b Compunds previously
investigated (20); EC3 values are included here for reference.
Results and Discussion
The aim of the present study was to investigate how the
removal of the endocyclic double bond of the previously
investigated allylic hydroperoxides 3-5 (20) influenced the
radical formation and sensitizing capacities. Thus, the alkylic
hydroperoxides 12, 14, and 17 have been synthesized and
subjected to radical-trapping experiments as well as sensitizing
studies. The radical-trapping experiments utilized Fe(III)TPPCl
as a radical initiator and TMIO as a radical trapper (Scheme
3). The results show that small changes in structure, like the
omission of the endocyclic double bond or the addition of a
methyl group, resulted in large differences in radical formation.
All of the investigated hydroperoxides formed large amounts
of radicals; however, the identity of the radicals was influenced
by the structure of the parent hydroperoxide. The results also
indicate that alkoxyl radicals seem to be more important than
carbon-centered radicals in the immunogenic complex formation.
The sensitizing capacities were investigated in the murine
LLNA, where all hydroperoxides were found to be potent
sensitizers.
Synthesis. Compound 10 was reacted with hydrazintosylate
to afford hydrazone 9 that was reduced with sodium borohydride
to hydrazine 11 (Scheme 2). Attempts to purify 11 using liquid
chromatography on silica gel resulted in decomposition. Because
of this instability, no satisfactory NMR spectra could be
obtained, and 11 was used without further purification. Treat-
ment with aqueous hydrogen peroxide yielded four diastereo-
mers of hydroperoxide 12, three of which could be enriched as
major products through consecutive preparative HPLC.
Extensive efforts to transform alcohol 7 (Table 2) directly to
hydroperoxide 14 via substitution were unsuccessful. In general,
no reaction was observed, or products formed and identified
originated from the elimination of water (4-isopropenyl-1,2-
dimethyl-cyclohexene) or addition to the isoprene unit, typically
in a Markovnikov manner. Further attempts to substitute the
1
major isomer: H NMR: δ 1.55-1.77 (m, 2H, H5), 1.66 (s, 3H,
H11), 2.15 (s, 6H, H1 and H8), 2.35-2.43 (m, 2H, H6), 2.43-2.57
(m, 2H, H3), 2.59-2.67 (m, 1H, H4), 4.74-4.84 (m, 2H, H10).
13C NMR: δ 18.7 (C11), 26.6 (C5), 30.3 (C1 or C8), 30.4 (C1 or
C8), 41.4 (C6), 42.1 (C4), 48.4 (C3), 112.9 (C10), 145.9 (C9), 207.9
(C2), 208.7 (C7). MS (API-ES, 70 eV) m/z (%): 205 [M + Na]
(64), 183 [M + H] (75), 165 (83), 123 (57), 107 (100). Anal. calcd
for C11H18O2: C, 72.49; H, 9.95. Found: C, 72.11; H, 10.18.
Compound 7 (10.1 mg, 3.8%), 1H and 13C NMR data agreed with
the literature (23).
Trapping Experiments with 17. Starting from 0.13 g (0.76 mmol)
of 17, two products were isolated using flash chromatography on
silica gel eluting with hexane/ethyl acetate (9:1). 3′-Isopropyl-1′-
(1,1,3,3-tetramethylisoindolyl-2-oxy)-heptan-6′-one (22, 39 mg,
1
15%): H NMR: δ 0.91 (d, J ) 7.0 Hz, 6H, H9′ and H10′), 1.46
(br s, 12H, H10 and H11 and H12 and H13), 1.45-1.57 (m, 3H,
HA2′ and HA4′ and H3′ or H8′), 1.60-1.80 (m, 3H, HB2′ and HB4′
and H3′ or H8′), 2.17 (s, 3H, H7′), 2.41-2.57 (m, 2H, H5′),
3.90-4.00 (m, 2H, H1′), 7.09-7.14 (m, 2H, H4 and H7), 7.21-7.24
(m, 2H, H5 and H6). 13C NMR: δ 19.0 (C9′ or C10′), 19.2 (C9′ or
C10′), 25.0 (C4′), 25.5 (2C, C10 and C13 or C11 and C12), 29.6
(C3′), 29.9 (C2′), 30.1 (C7′), 30.5 (2C, C10 and C13 or C11 and
C12), 40.6 (C8′), 42.2 (C5′), 67.2 (C1 and C3), 76.3 (C1′), 121.6
(C4 and C7), 127.3 (C5 and C6), 145.3 (C8 and C9), 209.4 (C6′).
MS (API-ES, 120 eV) m/z (%): 346 [M + H] (100), 160 (36).
Anal. calcd for C22H35NO2: C, 76.92; H, 9.68. Found: C, 77.02; H,
10.47. 4-Isopropenyl-1-methyl-cyclohexane-1-ol (18: major isomer,
1
16 mg, 13%; minor isomer, 14 mg, 11%), H and 13C NMR data
agreed with the synthesized reference.
Sensitization Experiments in Mice. The sensitizing potency of
the hydroperoxides was investigated using the LLNA (26) as
previously reported (20). All animal procedures were approved by
the local ethics committee. Each hydroperoxide was tested in five
different concentrations (Table 1), using mice in groups of three.

682 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Johansson et al.
Table 2. Relationship between Structure, Radical Formation, and Products Isolated in the Radical-Trapping Experiments with
Hydroperoxides 12, 14, and 17
a Secondary or tertiary hydroperoxide. b Oxygen-centered radicals formed by oxygen-oxygen bond cleavage (pathway i). c Carbon-centered radicals
formed by 1,2-shift (pathway ii). d n.d., not detected. e Primary or secondary carbon-centered radical formed by ꢀ-scission (pathway iii).
hydroxyl into a chloride, bromide, mesylate, or tosylate leaving
group were also unsuccessful, again resulting in elimination of
water or addition to the isoprene unit. Instead, the peroxyl
functionality was introduced as a silyl peroxide via a cobalt-
catalyzed addition to an alkene. The appropriate alkene 6 was
synthesized from 10 using standard Wittig conditions (Scheme
2). The cobalt catalyst [Co(thd)2] was prepared by treating
cobalt(II) chloride with 2,2,6,6-tetramethyl-3,5-heptanedione and
sodium hydroxide in water. Co(thd)2 precipitated and was
purified by sublimation to give purple crystals. Treatment of
alkene 6 with triethylsilane, tert-butyl hydroperoxide, and
Co(thd)2 under oxygen atmosphere furnished silyl peroxide 13
as well as the byproduct caused by peroxidation of the
isopropenyl group. Compound 13 was desilylated by treatment
with acidic methanol yielding hydroperoxide 14 (Scheme 2).
Attempted synthesis of hydroperoxide 23 (Figure 1) by silyl
peroxidation of the corresponding dialkene with the cobalt-
catalyzed method used for hydroperoxide 14 was abandoned

Structural Influence on Sensitizing Capacity
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 683
Scheme 3. Mechanistic Proposal That Accounts for the
Products Isolated in the Trapping Experimentsa
due to very low yields of the correct silyl peroxide. The
formation of the isopropenyl double bond via tosylate elimina-
tion was unsuccessful due to the incompatibility of the previ-
ously installed silyl peroxide with the elimination conditions.
In view of the proposed mechanism for immunogenic complex
formation of unsaturated hydroperoxides (27), it was interesting
to investigate the sensitizing capacity of a fully saturated
hydroperoxide. Furthermore, no reactions including the isopro-
penyl unit had been observed in the trapping experiments with
hydroperoxides 3-5 (20). Therefore, it was decided to synthe-
size the fully saturated analogue of hydroperoxide 23, that is,
hydroperoxide 17 (Figure 1). To do this, alkene 15 was
synthesized from ketone 16 using standard Wittig conditions
(Scheme 2). Compound 15 was then treated with triethylsilane,
tert-butyl hydroperoxide, and Co(thd)2 under oxygen atmosphere
yielding silyl peroxide 24 that was subsequently desilylated with
acidic methanol to furnish hydroperoxide 17.
Alcohol 18 (Table 2) was synthesized as a reference
compound starting from limonene oxide that was transformed
into ꢀ-terpineol using lithium aluminum hydride. ꢀ-Terpineol
was reduced to the fully saturated alcohol 18 by treatment with
palladium on carbon.
Sensitizing Capacity of the Hydroperoxides. The LLNA
was used to determine the sensitizing capacity of hydroperoxides
12, 14, and 17 (Figure 1). Each hydroperoxide was tested at
five different concentrations along with a control group that was
tested with the vehicle alone. Compound 12 had an EC3 value
of 0.065 M (1.1% w/v), 14 had an EC3 value of 0.037 M (0.68%
w/v), and 17 had an EC3 value of 0.020 M (0.35% w/v) (Table
1). Thus, 12, 14, and 17 are potent sensitizers, and the found
EC3 values are in agreement with the values of several other
previously tested hydroperoxides (8, 9, 14, 15, 20). This suggests
that the sensitizing capacities of the hydroperoxides are mainly
correlated to the hydroperoxide group and depend less on the
structure of the compound.
a Compound 12: R ) Me, R′ ) H, and R′′ ) isopropenyl in position 5.
Compound 14: R ) Me, R′ ) Me, and R′′ ) isopropenyl in position 5.
Compound 17: R ) H, R′ ) Me, and R′′ ) isopropyl in position 4.
Compound 20: n ) 1, R ) Me, and R′′ ) isopropenyl. Compound 22: n
) 2, R ) H, and R′′ ) isopropyl. aSee the specified scheme for mechanistic
details.
Chemical-Trapping Experiments with TMIO. Trapping
experiments with hydroperoxides 12, 14, and 17 were performed
to identify and quantify the major products formed. The
hydroperoxides were allowed to react in a 1:1 mixture of
acetonitrile/water using Fe(III)TPPCl as an initiator, cleaving
the oxygen-oxygen bond of the hydroperoxide group homolyti-
cally. TMIO was included in the experiments to trap carbon-
centered radicals. The outcome of the trapping experiments was
analyzed by NMR, GC/FID, and LC/MS.
or dioxygen and subsequent expulsion of TMIO-H or
hydrogen peroxide furnishing the carbonyl group in 10 (vide
infra, Scheme 4); and (iii) ꢀ-scission generating primary or
secondary carbon-centered alkyl radicals, followed by trap-
ping by TMIO yielding adducts 20 and 22 or by dioxygen
yielding a peroxyl radical that reacts further to give carbonyl
compound 21 (vide infra, Scheme 5). The preferred reaction
pathways together with formed radicals and products for each
hydroperoxide are summarized in Table 2.
The products isolated from the trapping experiment with
hydroperoxide 12 indicate a preference of the initially formed
alkoxyl radical (24) to undergo a 1,2-shift, eventually resulting
in 10 as the major product (Scheme 3). Alkoxyl radical 24 can
abstract a hydrogen atom and form 19 or undergo a 1,2-shift to
form the carbon-centered radical 27 (Scheme 4). Being a carbon-
centered radical, 27 is most likely trapped by TMIO, resulting
in adduct 28. However, adduct 28 was not detected. It is believed
that it decomposes to 10 and TMIO-H. The addition of dioxygen
to 27 yields peroxyl radical 29, which can abstract a hydrogen
atom and form 30. However, hydroperoxide 30 was not detected;
it is believed that it decomposes to 10 and H2O2 through a
general acid-catalyzed fragmentation. The yields of 10 and 19
are further affected by the actions of hydroxyl and hydroperoxyl
radicals on hydroperoxide 12 and alcohol 19 producing carbon-
centered radicals 27 and 31, which ultimately form 10.
Quantification of the products formed from hydroperoxides
12 and 14 was performed using GC/FID. Hydroperoxide 14,
TMIO, and the expected products 7, 10, and 19 (Table 2) were
calibrated against an internal standard. Hydroperoxide 12 partly
degraded to 19 on the GC column and was therefore not
quantified. However, no traces of unreacted hydroperoxide 12
were found in the reaction mixture, according to NMR and LC/
MS analyses.
The products identified in the trapping experiment with 12
were 19 (3.2%) and 10 (41%; Table 2). The products identified
in the trapping experiment with 14 were alcohol 7 (3.8%),
diketone 21 (14%), and TMIO adduct 20 (47%). The products
identified in the trapping experiment with 17 were alcohol 18
(24%) and TMIO adduct 22 (15%).
The homolytic cleavage of the oxygen-oxygen bond yields
alkoxyl radicals (24-26) corresponding to each hydroper-
oxide. The alkoxyl radicals react according to three major
pathways (Scheme 3): (i) hydrogen abstraction yielding the
analogous alcohols; (ii) a 1,2-shift generating a carbon-
centered R-hydroxyl radical, followed by trapping by TMIO
The products isolated from the trapping experiments with
hydroperoxide 14 clearly indicate a preference for alkoxyl
radical 25 reacting according to the ꢀ-scission pathway, resulting

684 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Johansson et al.
Scheme 4. Mechanistic Proposal for the Formation of 10 in the Trapping Experiment with Hydroperoxide 12
in TMIO adduct 20 and diketone 21 as the major products.
Alkoxyl radical 25 is formed by homolytic cleavage of 14. It
can abstract a hydrogen atom and form alcohol 7 or react via a
ꢀ-scission pathway (Scheme 3). Three ꢀ-scissions are possible
for radical 25. ꢀ-Scission of the methyl group would generate
10; however, no evidence for the formation of 10 was found.
ꢀ-Scission of one of the bonds in the cyclohexane ring would
result in ring opening and generate either a primary or a
secondary carbon-centered radical. No TMIO adduct corre-
sponding to the primary carbon-centered radical was detected.
The preferred ꢀ-scission generates the secondary carbon radical
32, which is trapped by TMIO yielding adduct 20 (Scheme 5).
The addition of dioxygen to 32 yields peroxyl radical 33, which
abstracts a hydrogen atom to form hydroperoxide 34 or
dimerizes to tetroxide 35. Hydroperoxide 34 can form diketone
21 via acid-catalyzed loss of water, whereas dimer 35 decom-
poses into diketone 21 and alcohol 36 via the Russell mechanism
(28). Alcohol 36 could potentially be transformed into ketone
21 by hydrogen abstraction followed by trapping, either by
TMIO or dioxygen. This would be followed by an expulsion
of TMIO-H or hydrogen peroxide, respectively, forming dike-
tone 21. As alcohol 36 is not detected, dimer 35 is either not
formed, or 36 is consumed in the reaction.
indicate that hydrogen abstraction is the favored reaction
pathway in the trapping experiment with 17.
Structural Analysis of Carbon Radical Adducts and
Nonradical Products from the Trapping Experiments. The
TMIO adducts and nonradical products from the radical-trapping
experiments were identified using a combination of NMR and
LC/MS. Liquid chromatography confirmed the purity of the
samples, and mass spectroscopy identified ion masses corre-
sponding to [M + H]+ or [M + Na]+. In the case of the TMIO
adducts, fragmentation patterns indicated the presence of TMIO
as well as a terpene residue. NMR shifts for individual 1H and
13C signals corresponded to previously characterized adducts
and known structural elements of the present hydroperoxides.
The structures of the terpene residues and the nonradical
products were elucidated using a combination of 1D and 2D
NMR experiments.
The structure of TMIO adduct 20 (Figure 2) was confirmed
by 2D NMR correlations. The integrity of the TMIO unit was
2
established by J(C, H) correlations between the hydrogens in
positions 10-13 and the quaternary carbons in positions 1, 3,
3
2
3
8, and 9 as well as J(H, H), J(C, H), and J(C, H) couplings
between the aromatic hydrogens in positions 4-7 and the
aromatic carbons in positions 4-9. The preservation of the
isopropenyl unit was confirmed by 4J(H, H) and 3J(C, H)
correlations between the hydrogens in position 10′ and the
hydrogens and the carbon, respectively, in position 11′. Fur-
Homolytic cleavage of hydroperoxide 17 generates alkoxyl
radical 26 that can abstract a hydrogen atom and form alcohol
18 (Table 2) or undergo ꢀ-scission, which results in opening of
the cyclohexane ring. There are two possible ring-opening
ꢀ-scissions, both resulting in carbon-centered radical 37, which
is subsequently trapped by TMIO, forming adduct 22 (Scheme
5). ꢀ-Scission of the methyl group would generate 16 (Scheme
2), but no evidence for ketone formation was found. The
isolation and identification of alcohol 18 as the major product
2
thermore, the hydrogens in position 10′ displayed a J(C, H)
coupling to the carbon in position 9′. Both carbon and hydrogens
3
in position 10′ displayed J(C, H) couplings to the carbon and
the hydrogen in position 4′; this confirms the attachment of the
isopropenyl group to the alkane chain. The carbonyl group was
2
identified by its characteristic carbon shift and the J(C, H)

Structural Influence on Sensitizing Capacity
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 685
Scheme 5. Mechanistic Proposal for the Formation of Ketone 21 in the Trapping Experiment with Hydroperoxide 14a
a Compound 20: n ) 1, R ) Me, and R′′ ) isopropenyl. Compound 22: n ) 2, R ) H, and R′′ ) isopropyl. Compound 32: n ) 1, R ) Me, and R′′ )
isopropenyl. Compound 26: n ) 2, R ) H, and R′′ ) isopropyl.
in position 4. The carbonyl groups and the acetyl structural
fragments were identified by their characteristic carbon shifts
and the 2J(C, H) couplings between the hydrogens in positions
1 and 8 to the carbons in positions 2 and 7, respectively. The
attachment of the carbonyls to the alkane chain was evident
2
from the J(C, H) couplings to the hydrogens in positions 3
and 6, respectively.
TMIO adduct 22 (Figure 2) was identified by 2D NMR
Figure 2. Products derived from carbon-centered radicals and isolated
in the radical-trapping experiments.
correlations. The preservation of the TMIO unit was confirmed
by the J(C, H) couplings between the hydrogens in positions
couplings to the adjacent hydrogens in positions 1′ and 3′. The
2
attachment and position of TMIO were identified by the typical
10-13 and the carbons in positions 1 and 3, as well as the 2J(C,
H) and 3J(C, H) couplings of the aromatic hydrogens in positions
4-7 to the quaternary carbons in positions 8 and 9. The
connection of the TMIO unit to the alkane chain was confirmed
by the characteristic shift of the carbon in position 1′. The
integrity of the isopropyl group was confirmed by the 3J(C, H)
couplings of the hydrogens in positions 9′ and 10′ to the carbon
in position 8′. The attachment and position of the isopropyl
group to the alkane chain were evident from the 2J(C, H)
coupling between the carbon in position 8′ and the hydrogen in
position 3′. The identity and position of the acetyl group were
confirmed by the typical shift of the carbonyl carbon and the
2J(C, H) couplings of the carbonyl carbon to the hydrogens in
positions 5′ and 7′, respectively.
2
shift of the carbon in position 7′ and the J(C, H) correlation
between the hydrogens in position 8′ and the carbon in position
3
7′ as well as the J(H, H) couplings between the hydrogens in
positions 6′ and 8′ with the hydrogen in position 7′.
The identity of product 21 (Figure 2) was supported by 2D
NMR correlations. The integrity of the isopropenyl unit was
4
3
confirmed by J(H, H) and J(C, H) correlations between the
hydrogens in position 10 and the hydrogens and the carbon, in
3
position 11. J(C, H) correlations between the hydrogens in
position 10 and the carbon in position 4 as well as between the
carbon in position 11 and the hydrogen in position 4 confirm
the attachment of the isopropenyl group to the alkane chain.
The position of the isopropenyl group on the alkane chain was
3
identified by the J(H, H) couplings between the hydrogens in
In the 13C NMR experiment with TMIO adduct 22 at room
position 5 and positions 4 and 6, respectively, and the 2J(C, H)
coupling connecting the hydrogens in position 3 to the carbon
temperature, no signals corresponding to the methyl groups in

686 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Johansson et al.
positions 10-13 were detected. To investigate if this was due
to inversion of the nitrogen configuration, a new set of
experiments was run at lowered temperatures. At -50 °C, two
new signals were displayed; at -10 °C, these signals were
broadened; at 10 °C, they were very broad; and at 25 °C, they
were missing. This confirmed our hypothesis.
sitizers (31, 32). Thus, the formation of these compounds cannot
account for the high sensitizing capacity of hydroperoxide 12.
This implies the importance of radicals in immunogenic complex
formation of hydroperoxides. As only small amounts of alcohol
and no TMIO adducts are detected in the trapping experiments,
the alkoxyl radicals are likely to be important in the im-
munogenic complex formation.
Radical Formation Pathways. The structures of the alkylic
hydroperoxides clearly influence which reaction pathways will
dominate the radical formation and thereby the products
identified in the radical-trapping experiments (Table 2). In our
previous study on the corresponding allylic hydroperoxides (20),
products identified were attributed to three reaction pathways:
hydrogen abstraction, 1,2-shift, and 1,3-cyclization (Scheme 1).
In the present study, products identified can be attributed to
hydrogen abstraction (alcohols 7, 18, and 19, Scheme 3) and
1,2-shift (10) but not to the 1,3-cyclization. Instead, products
formed by ꢀ-scission (TMIO adducts 20 and 22 and ketone 21)
were identified. The lack of cyclization products is explained
by the lack of the endocyclic double bond. The high yields of
10 and ꢀ-scission products compared with the yields of alcohols
can be explained by the rate difference between intra- and
intermolecular reactions. There are two possible intramolecular
reaction mechanisms: 1,2-shift and ꢀ-scission. The 1,2-shift
requires a hydrogen on the hydroperoxide-bearing carbon. Thus,
no products from this pathway were detected from the tertiary
hydroperoxides; instead, ꢀ-scission products were identified. The
lower yield of alcohol and higher yield of ꢀ-scission products
in the trapping experiment with 14, as compared with the
trapping experiment with 17, indicate that the secondary carbon-
centered radical originating from 14 is more easily formed than
the primary carbon-centered radical originating from 17. This
is in accordance with the general stability of carbon-centered
radicals.
In the radical-trapping experiment with 14, the high yields
of 20 and 21 indicate the formation of large amounts of carbon-
centered radicals (Table 2). Only a small amount of the
corresponding alcohol is formed, indicating that only a small
amount of alkoxyl are radicals available for the formation of
immunogenic complexes. As both 12 and 14 are potent
sensitizers, carbon-centered radicals may be less important than
oxygen-centered radicals in the immunogenic complex forma-
tion. Ketone 21 is unlikely to be responsible for the sensitizing
capacity of hydroperoxide 14. The reactive groups of 21 are
two carbonyl groups. The sensitizing potential of a carbonyl
group is dependent on the electron deficiency of the carbony1
atom (33). In analogy with 10, the substituents on both of the
carbonyl groups of ketone 21 are electron-donating alkyl groups;
this makes it unlikely that the electron deficiency of the carbonyl
carbons is pronounced enough to generate a potent sensitizer.
In the radical-trapping experiment with 17, a large amount
of alcohol is formed together with a somewhat lower amount
of TMIO adducts (Table 2). Thus, a large amount of alkoxyl
radicals are potentially available for immunogenic complex
formation, together with a slightly lower amount of carbon-
centered radicals. In our previous studies of allylic hydroper-
oxides we found that hydroperoxide 4 had a significantly higher
sensitizing capacity than hydroperoxides 3 and 5 (34). These
results were obtained using a modified LLNA comprising
nonpooled lymph nodes and statistical evaluation. A similar
study has not been performed on hydroperoxides 12, 14, and
17. However, there are major similarities in the radical formation
and the sensitizing capacities of hydroperoxides 4 and 17. Thus,
it is possible that hydroperoxide 17 has a higher sensitizing
capacity than hydroperoxides 12 and 14. If so, this would be
an indication of both the importance of the amount of radicals
formed and the potency of the alkoxyl radicals.
In summary, the structure of the hydroperoxides and the rates
of the available reaction pathways will combine to dictate the
identity and quantity of the radicals formed (Table 2). These
radicals are potentially available for formation of immunogenic
hapten-protein complexes.
Immunogenic Complex Formation. The exact mechanism
of immunogenic complex formation of hydroperoxides is not
known. It is postulated that it takes place via a radical
mechanism (6-9), but only one specific mechanism have been
proposed (27). As this mechanism only accounts for olefinic
hydroperoxides, it is likely that other radical mechanisms are
active in parallel. Their individual activity will depend on the
identity of the formed radicals. The activity of different radical
mechanisms is likely to influence the sensitizing capacity of
the parent hydroperoxide. Thus, the identities of the formed
radicals are important for the sensitizing capacity of the
hydroperoxide. Furthermore, there is a threshold for the im-
munogenic response (29), meaning that a minimum number of
immunogenic complexes need to be formed to trigger the
response. This makes the amounts of radicals formed important,
as more radicals potentially mean a higher amount of im-
munogenic complexes. In the radical-trapping experiments, the
amount of alcohol indicates the amount of alkoxyl radicals
available for immunogenic complex formation, whereas the
amount of TMIO adducts indicate the amount of carbon-centered
radicals available for immunogenic complex formation.
We have recently presented results that suggest the thiol-ene
radical reaction as a possible mechanism for the formation of
immunogenic hapten-protein complexes from olefinic hydro-
peroxides (27). In the proposed mechanism, a thiyl radical is
formed from the thiol group of cysteine, probably by hydrogen
abstraction. Thereafter, the thiyl radical adds to double bonds
of unsaturated compounds derived from the hydroperoxide. We
have identified products resulting from the addition of the thiol
group of cysteine and GSH over the double bonds of carvone
and carveol. These products were formed in a system that
initially contained only 3, Fe(III)TPPCl, and cysteine or GSH.
No adducts were detected when Fe(III)TPPCl was omitted from
the reaction mixture, indicating a radical mechanism and the
importance of radical formation.
Numerous radicals are formed in the proposed mechanisms
of this work as well as in our previous investigation (20). Any
of these radicals could potentially abstract a hydrogen atom from
the thiol group of cysteine, thereby initiating the thiol-ene
reaction. This increases the number of possible initiations, while
simultaneously being integrated in the formation of nonradical
unsaturated compounds (except for 17, vide infra). Thus, the
thiol-ene reaction is a possible mechanism for the formation
of immunogenic hapten-protein complexes of hydroperoxides
12 and 14.
Hydroperoxide 12 does not form any TMIO adducts in the
radical-trapping experiment. Instead, large amounts of 10 and
minor amounts of 19 (Table 2) were identified. Compound 10
is a known nonsensitizer (30), and 19 is very unlikely to be a
sensitizer as other aliphatic alcohols tested have been nonsen-

Structural Influence on Sensitizing Capacity
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 687
The formation of thiyl radicals is a key step in the proposed
thiol-ene mechanism. The thiyl radicals are probably formed
by hydrogen abstraction, which might explain why the alkoxyl
radicals seem to have a higher impact on the sensitizing capacity
than the carbon-centered radicals. On the basis of bond
dissociation energies, the formation of an oxygen-hydrogen
bond is energetically favored over the formation of a carbon-
hydrogen bond (35). This might result in more thiyl radicals
formed for a given amount of alkoxyl radicals as compared with
the same amount of carbon-centered radicals.
Being fully saturated, hydroperoxide 17 cannot form im-
munogenic complexes via the thiol-ene mechanism proposed
for olefinic hydroperoxides. However, several other radical
mechanisms might still be possible, for example, the direct
addition of alkoxyl radicals to amino acid side chains or amino
acid radicals. Hydrogen abstraction from fully saturated positions
on compounds derived from 17 would give carbon-centered
radicals that could possibly react with amino acid radicals in
termination reactions. The high sensitizing capacity of 17,
together with its incompatibility with the thiol-ene reaction
mechanism, strongly indicates that several radical reaction
pathways are available for the radical immunogenic complex
formation of hydroperoxides.
If the alkyl groups attached to the hydroperoxide-bearing
carbon of a hydroperoxide are strongly electron donating and
have the properties necessary to stabilize a plus charge on that
carbon, a carbocation can be formed from the hydroperoxide
under acidic conditions (36, 37). This carbocation would be
available for SN1 attack by amino acid side chains. However,
no reports of alkylic hydroperoxides reacting in this fashion
have been found in the literature. Several nonradical products
were formed from the hydroperoxides in the radical-trapping
experiments in both this and our previous study of allylic
hydroperoxides (20). In theory, these products could be attacked
by amino acid side chains in nucleophilic-electrophilic reac-
tions. Nevertheless, none of these products have a known or
suspected sensitizing capacity comparable with the hydroper-
oxides investigated and are therefore unlikely to be responsible
for the high sensitizing capacity of the hydroperoxides. Inde-
pendent of structure, all of the hydroperoxides in this as well
as in our previous study (20) form large amounts of radicals
and are potent sensitizers. Taken together, we find strong
indications of a radical mechanism for the immunogenic
hapten-protein complex formation of hydroperoxides.
All hydroperoxides tested have been potent sensitizers with
small differences in their sensitizing capacities (see refs 8, 9,
14, 15, and 20 and this work). This may be explained by the
depletion of local antioxidant reserves by the formation of high
amounts of radicals in the skin (38, 39). Furthermore, this
depletion may facilitate the formation of immunogenic com-
plexes via a radical mechanism (27). This offers an explanation
to why all hydroperoxides are strong sensitizers and have the
ability to form specific antigens (11).
antioxidant defenses that is caused by the formation of high
amounts of radicals in the skin (38, 39). This mode of action is
compatible with the formation of specific antigens of hydrop-
eroxides (11). For two of the hydroperoxides investigated in
the present study, the formation of radicals and the simultaneous
presence of unsaturated terpenoids make the recently suggested
thiol-ene reaction (27) a possible mechanism for the formation
of immunogenic complexes. The thiol-ene mechanism is not
possible for the fully saturated hydroperoxide that is also
investigated in this study. This hydroperoxide is a potent
sensitizer and forms large amounts of radicals, results that
strongly indicate that more than one radical reaction pathway
for immunogenic complex formation of hydroperoxides is active.
Acknowledgment. We thank Professor Ann-Therese
Karlberg for valuable discussions and comments to the manu-
script, Assistant Professor Mate Erdelyi for valuable assistance
with NMR experiments, and Ph.D. Timothy M. Altamore as
well as Ph.D. Fredrik Lindahl for skillful laboratory assistance.
The work was performed within the Go¨teborg Science Centre
for Molecular Skin Research.
Supporting Information Available: 1H and 13C NMR data
for the minor isomers of (5R)-5-isopropenyl-2-methyl-cyclo-
hexane-1-hydroperoxide (12b-d) and full characterization data
for the minor isomer of 18. This material is available free of
charge via the Internet at http://pubs.acs.org.
References
(1) Nielsen, N. H., Linneberg, A., Menne´, T., Madsen, F., Fro¨lund, L.,
Dirksen, A., and Jo¨rgensen, T. (2001) Allergic contact sensitization
in an adult danish population: Two cross-sectional surveys eight years
apart (the Copenhagen allergy study). Acta Derm.-Venereol. 81, 31–
34.
(2) Rustemeyer, T., van Hoogstraten, I. M. V., von Blomberg, B. M. E.,
and Scheper, R. J. (2006) Mechanisms in allergic contact dermatitis.
In Contact Dermatitis (Frosch, P. J., Menne´, T., and Lepoittevin, J. P.,
Eds.) pp 11-43, Springer-Verlag, Berlin, Germany.
(3) Karlberg, A. T., Bergstro¨m, M. A., Bo¨rje, A., Luthman, K., and
Nilsson, J. L. G. (2008) Allergic contact dermatitissFormation,
structural requirements, and reactivity of skin sensitizers. Chem. Res.
Toxicol. 21, 53–69.
(4) Schmidt, R. J., Khan, L., and Chung, L. Y. (1990) Are free radicals
and not quinones the haptenic species derived from urushiols and other
contact allergenic mono- and dihydric alkylbenzenes? The significance
of NADH, glutathione, and redox cycling in the skin. Arch. Dermatol.
Res. 282, 56–64.
(5) Xia, Z. Y., Miyakoshi, T., and Yoshida, T. (2004) Lipoxygenase-
catalyzed polymerization of phenolic lipids suggests a new mechanism
for allergic contact dermatitis induced by urushiol and its analogs.
Biochem. Biophys. Res. Commun. 315, 704–709.
(6) Ga¨fvert, E., Shao, L. P., Karlberg, A. T., Nilsson, U., and Nilsson,
J. L. G. (1994) Contact allergy to resin acid hydroperoxidessHapten
binding via free-radicals and epoxides. Chem. Res. Toxicol. 7, 260–
266.
(7) Lepoittevin, J. P., and Karlberg, A. T. (1994) Interactions of allergenic
hydroperoxides with proteins: A radical mechanism? Chem. Res.
Toxicol. 7, 130–133.
(8) Mutterer, V., Gime´nez-Arnau, E., Karlberg, A. T., and Lepoittevin,
J. P. (2000) Synthesis and allergenic potential of a 15-hydroperoxya-
bietic acid-like model: Trapping of radical intermediates. Chem. Res.
Toxicol. 13, 1028–1036.
Conclusion
(9) Bezard, M., Karlberg, A. T., Montelius, J., and Lepoittevin, J. P. (1997)
Skin sensitization to linalyl hydroperoxide: Support for radical
intermediates. Chem. Res. Toxicol. 10, 987–993.
(10) Redeby, T., Nilsson, U., Altamore, T. M., Ilag, L., Ambrosi, A., Broo,
K., Bo¨rje, A., and Karlberg, A. T. (2010) Specific adducts formed
through a radical reaction between peptides and contact allergenic
hydroperoxides. Chem. Res. Toxicol. 23, 203–210.
(11) Christensson, J. B., Matura, M., Ba¨cktorp, C., Bo¨rje, A., Nilsson,
J. L. G., and Karlberg, A. T. (2006) Hydroperoxides form specific
antigens in contact allergy. Contact Dermatitis 55, 230–237.
(12) Gime´nez-Arnau, E., Haberkorn, L., Grossi, L., and Lepoittevin, J. P.
(2002) Identification of alkyl radicals derived from an allergenic cyclic
We conclude that small changes in structure, like the omission
of the endocyclic double bond or the addition of a methyl group,
result in large differences in radical formation of limonene
hydroperoxides and their analogues. Furthermore, our results
indicate that alkoxyl radicals seem to be more important in the
immunogenic complex formation than carbon-centered radicals.
All hydroperoxides investigated formed high amounts of radicals
and were potent sensitizers. A possible explanation for the high
sensitizing capacities of hydroperoxides is the weakening of the

688 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Johansson et al.
tertiary allylic hydroperoxide by combined use of radical trapping and
ESR studies. Tetrahedron 58, 5535–5545.
(13) Nilsson, U., Bergh, M., Shao, L. P., and Karlberg, A. T. (1996)
Analysis of contact allergenic compounds in oxidized d-limonene.
Chromatographia 42, 199–205.
immunogenic complexes via a radical pathway: A key step in allergic
contact dermatitis to olefinic hydroperoxides. Chem. Res. Toxicol. 22,
1774–1781.
(28) Russell, G. A. (1957) Deuterium-isotope effects in the autoxidation
of aralkyl hydrocarbons. Mechanism of the interaction of peroxy
radicals. J. Am. Chem. Soc. 79, 3871–3877.
(29) Basketter, D. A., Evans, P., Gerberick, G. F., and Kimber, I. A. N.
(2002) Factors affecting thresholds in allergic contact dermatitis: Safety
and regulatory considerations. Contact Dermatitis 47, 1–6.
(30) Nilsson, A. M., Ga¨fvert, E., Salvador, L., Luthman, K., Bruze, M.,
Gruvberger, B., Nilsson, J. L., and Karlberg, A. T. (2001) Mechanism
of the antigen formation of carvone and related alpha, beta-unsaturated
ketones. Contact Dermatitis 44, 347–356.
(31) Basketter, D. A., Gerberick, G. F., and Kimber, I. (1998) Strategies
for identifying false positive responses in predictive skin sensitization
tests. Food Chem. Toxicol. 36, 327–333.
(32) Ryan, C. A., Gerberick, G. F., Cruse, L. W., Basketter, D. A., Lea,
L., Blaikie, L., Dearman, R. J., Warbrick, E. V., and Kimber, I. (2000)
Activity of human contact allergens in the murine local lymph node
assay. Contact Dermatitis 43, 95–102.
(33) Aptula, A. O., Patlewicz, G., and Roberts, D. W. (2005) Skin
sensitization: Reaction mechanistic applicability domains for structure-
activity relationships. Chem. Res. Toxicol. 18, 1420–1426.
(34) Christensson, J. B., Johansson, S. G. H., Hagvall, L., Jonsson, C.,
Bo¨rje, A., and Karlberg, A.-T. (2008) Limonene hydroperoxide
analogues differ in allergenic activity. Contact Dermatitis 59, 344–
352.
(35) McMurry, J. (2004) Bond dissociation enthalpies table. Organic
Chemistry, p 154, Brooks/ColesThomson Learning, Belmont.
(36) Davies, A. G., Foster, R. V., and Nery, R. (1954) Organic peroxides.
Part IV. Alkyl oxygen heterolysis in organic peroxides. J. Chem. Soc.
2204–2209.
(37) Taljaard, B., Goosen, A., and Mccleland, C. W. (1987) Synthesis of
hydrogen peroxide: Acid-catalyzed decomposition of 9-hydroperoxy-
9-phenylxanthene and its derivatives. S. Afr. J. Chem. 40, 139–145.
(38) Shvedova, A. A., Kisin, E. R., Murray, A. R., Kommineni, C.,
Castranova, V., Mason, R. P., Kadiiska, M. B., and Gunther, M. R.
(2002) Antioxidant balance and free radical generation in vitamin
E-deficient mice after dermal exposure to cumene hydroperoxide.
Chem. Res. Toxicol. 15, 1451–1459.
(39) Murray, A. R., Kisin, E. R., Kommineni, C., Vallyathan, V.,
Castranova, V., and Shvedova, A. A. (2007) Pro/antioxidant status
and AP-1 transcription factor in murine skin following topical exposure
to cumene hydroperoxide. Carcinogenesis 28, 1582–1588.
(40) Basketter, D. A., Lea, L. J., Dickens, A., Briggs, D., Pate, I., Dearman,
R. J., and Kimber, I. (1999) A comparison of statistical approaches to
the derivation of EC3 values from local lymph node assay dose
responses. J. Appl. Toxicol. 19, 261–266.
(14) Sko¨ld, M., Bo¨rje, A., Harambasic, E., and Karlberg, A. T. (2004)
Contact allergens formed on air exposure of linalool. Identification
and quantification of primary and secondary oxidation products and
the effect on skin sensitization. Chem. Res. Toxicol. 17, 1697–1705.
(15) Hagvall, L., Ba¨cktorp, C., Svensson, S., Nyman, G., Bo¨rje, A., and
Karlberg, A. T. (2007) Fragrance compound geraniol forms contact
allergens on air exposure. Chem. Res. Toxicol. 20, 807–814.
(16) Sko¨ld, M., Hagvall, L., and Karlberg, A. T. (2008) Autoxidation of
linalyl acetate, the main component of lavender oil, creates potent
contact allergens. Contact Dermatitis 58, 9–14.
(17) Karlberg, A. T., Shao, L. P., Nilsson, U., Ga¨fvert, E., and Nilsson,
J. L. G. (1994) Hydroperoxides in oxidized D-limonene identified as
potent contact allergens. Arch. Dermatol. Res. 286, 97–103.
(18) Sko¨ld, M., Bo¨rje, A., Matura, M., and Karlberg, A. T. (2002) Studies
on the autoxidation and sensitizing capacity of the fragrance chemical
linalool, identifying a linalool hydroperoxide. Contact Dermatitis 46,
267–272.
(19) Groot, A. C., and Frosch, P. J. (1997) Adverse reactions to fragrances.
A clinical review. Contact Dermatitis 36, 57–86.
(20) Johansson, S. G. H., Gime´nez-Arnau, E., Grøtli, M., Karlberg, A.-T.,
and Bo¨rje, A. (2008) Carbon- and oxygen-centered radicals are equally
important haptens of allylic hydroperoxides in allergic contact der-
matitis. Chem. Res. Toxicol. 21, 1536–1547.
(21) Griffiths, P. G., Moad, G., Rizzardo, E., and Solomon, D. H. (1983)
Synthesis of the radical scavenger 1,1,3,3-tetramethylisoindolin-2-
yloxyl. Aust. J. Chem. 36, 397–401.
(22) Bergstro¨m, M. A., Luthman, K., Nilsson, J. L. G., and Karlberg, A. T.
(2006) Conjugated dienes as prohaptens in contact allergy: In vivo
and in vitro studies of structure-activity relationships, sensitizing
capacity, and metabolic activation. Chem. Res. Toxicol. 19, 760–769.
(23) Pyun, H. J., Coates, R. M., Wagschal, K. C., McGeady, P., and
Croteau, R. B. (1993) Regiospecificity and isotope effects associated
with the methyl-methylene eliminations in the enzyme-catalyzed
biosynthesis of (R)- and (S)-limonene. J. Org. Chem. 58, 3998–4009.
(24) Gurudutt, K. N., Rao, S., and Srinivas, P. (1992) Stereospecific
synthesis of trans-beta terpineol. FlaVour Fragrance J. 7, 343–345.
(25) Hammond, G. S., Wu, C. H. S., and Nonhebel, D. C. (1963) Chelates
of ꢀ-diketones. V. Preparation and properties of chelates containing
sterically hindered ligands. Inorg. Chem. 2, 73–76.
(26) Kimber, I., Hilton, J., Dearman, R. J., Gerberick, G. F., Ryan, C. A.,
Basketter, D. A., Scholes, E. W., Ladics, G. S., Loveless, S. E., House,
R. V., and Guy, A. (1995) An international evaluation of the murine
local lymph node assay and comparison of modified procedures.
Toxicology 103, 63–73.
(27) Johansson, S. G. H., Redeby, T., Altamore, T. M., Nilsson, U., and
Bo¨rje, A. (2009) Mechanistic proposal for the formation of specific
TX900433N