QSAR for the cytotoxicity of 2-alkyl or 2,6-dialkyl, 4-X-phenols: The nature of the radical reaction

Article abstract of DOI:10.1039/b201478e

In a continuation of studies on the radical mediated toxicity of phenols to leukemia cells, a set of di- and tri-substituted phenols with mostly alkyl substituents in the ortho position were examined. These analogs are similar in structure to the commercial antioxidants BHA and BHT. A QSAR analysis of their growth inhibitory constants led to the formulation of this simple but unusual equation based on 18 congeners: log 1/C = -0.47E_s-2 + 2.42R_R + 2.43; r³ = 0.934 E_S-2 is the Taft steric parameter for the larger of the two ortho substituents while E_R is Otsu's radical parameter, which was originally defined to correlate radical reactions.

Full text of DOI:10.1039/b201478e

View Article Online / Journal Homepage / Table of Contents for this issue
QSAR for the cytotoxicity of 2-alkyl or 2,6-dialkyl, 4-X-phenols:
the nature of the radical reaction
a
a
a
b
Cynthia D. Selassie,* Rajeshwar P. Verma, Sanjay Kapur, Alan J. Shusterman and
Corwin Hansch
a
2
a
Chemistry Department, Pomona College, Claremont, CA 91711, USA.
E-mail: cselassie@pomona.edu; Fax: 909 607-7726; Tel: 909 621-8446
Chemistry Department, Reed College, Portland, OR 97202, USA
b
Received (in Cambridge, UK) 11th February 2002, Accepted 11th April 2002
First published as an Advance Article on the web 3rd May 2002
In a continuation of studies on the radical mediated toxicity of phenols to leukemia cells, a set of di- and tri-
substituted phenols with mostly alkyl substituents in the ortho position were examined. These analogs are similar in
structure to the commercial antioxidants BHA and BHT. A QSAR analysis of their growth inhibitory constants led
to the formulation of this simple but unusual equation based on 18 congeners:
ES-2 is the Taft steric parameter for the larger of the two ortho substituents while E is Otsu’s radical parameter, which
R
was originally defined to correlate radical reactions.
Introduction
electron donor substituents. Recent analyses of substituent
effects on phenoxyl radical formation have shown that electron
In the last ten years, the phenolic moiety has come under
increased scrutiny because of its bewildering array of biological
activities in vitro and in vivo. It appears to act as a potent
donating substituents such as OH and NH have lower BDEs
2
than predicted, indicating that they can stabilize the radical via
1
6,7
spin delocalization. Using ab initio and density functional
antioxidant in flavonoids yet it wields considerable toxicity as
theory (DFT), Brinck et al. have shown that the ∆BDEs of
2–4
in the case of p-cresol and diethylstilbestrol. This unusual
behavior of phenols has been attributed to its univalent oxid-
ation to form aryloxyl free radicals. It has also led to numerous
comprehensive studies on its biological activity at the molecular
and cellular level.
In a recent study on the cytotoxicity of a series of simple and
phenols with electron-withdrawing substituents are mostly7
determined by the polar stabilization of the parent molecule.
The polar effect is generally reduced for the electron-donating
phenols and is also found to destabilize the phenol. In these
cases, the spin delocalization effect predominates and leads to
radical stabilization. Radical effects are smaller and more
irregular for electron withdrawing substituents. The success of
5
complex mono-substituted phenols versus L1210 leukemia cells,
the following QSAR 1 was obtained for electron-rich phenols.
5
ϩ
correlations between BDE values and σ values is consistent
with these conclusions. In a similar study, Tomiyama et al.
demonstrated by a comparison of enthalpy (∆H ) and activ-
Inhibition of growth of L1210 cells by phenols with electron
releasing substituents
ation energies (E ) that electron-donating substituents lower
a
these energies and cause an increase in antioxidant activities
whereas electron-withdrawing substituents lead to an increase
in the transition state barrier and a decrease in antioxidant
(
1)
8
activity. A recent extensive survey of radical reactions reveals
ϩ
that Brown’s σ parameter is useful in delineating the role of
In eqn. (1), BDE represents the homolytic O–H bond dis-
sociation energy that strongly implicates a radical mechanism
of toxicity, while C constitutes the molar concentration of
radicals in various reactions; numerous examples of phenoxyl
radical formation in chemical as well as biological systems are
well correlated by σ . Radical formation of di-and poly-
ϩ 9
X-phenol that induces 50% inhibition of growth in the cell line
substituted phenols is also well defined by a summation of the
σ values.
2
ϩ
(
n is the number in the sample, r the correlation coefficient,
2
2
s the standard deviation and q the cross validated r ). The BDE
term can be effectively replaced by σ as in QSAR 2, but the
correlation is less robust (r = 0.895). This may be attributed in
Electron withdrawing substituents on the phenol ring also
induced toxicity, albeit much less than their electron donating
counterparts. The QSAR developed was considerably different
ϩ
2
ϩ
10
part to the lack of accurate σ values for complex, estrogenic
from that of QSAR 1 and 2 as see in QSAR 3.
phenols such as estradiol, estriol, equilin, equilenin, diethyl-
stilbesterol and bisphenol A which required estimation.
Inhibition of growth of L1210 cells by phenols with electron
withdrawing substituents
(
2)
(
3)
ϩ
The fact that σP cytotoxicity parallels the BDE toxicity can
be interpreted in terms of the large radical stabilizing effects of
In eqn. (3), C represents the concentration of X-phenol that
inhibits cell growth by 50%. Mono-ortho substituted phenols
1
112
J. Chem. Soc., Perkin Trans. 2, 2002, 1112–1117
This journal is © The Royal Society of Chemistry 2002
DOI: 10.1039/b201478e

View Article Online
5
were fitted well by eqn. (1) and (2). When sequestered alone,
of ES-2 for the 2 and 6 positions did not yield a significant
ϩ
these positional isomers showed a strong dependence on σP
QSAR. There also appears to be no steric effect wielded by
ϩ
and BDE but lacked any significant dependence on hydro-
phobic or steric parameters, suggesting that the formation of
the ortho substituted aryloxyl radical may be mediated by a
simple unencumbered radical or the orientation of the hydroxy
groups may be such that the hydroxy group points “away” from
the substituent and thus no repulsive interaction between the
4-substituents. Using σ in place of ER in the above QSAR
2
yields a poorer result (n = 18, r = 0.797, ρ = 0.54). If the 2-Me,
2
4-OMe congener is removed (n = 17, r = 0.862, ρ = 0.73), the
correlation is slightly better but the sign of the rho (ρ) value
remains positive. The positive ρ associated with QSAR 6 is
unusual and in stark contrast to QSAR 1. Otsu’s E parameter
R
11
two groups is seen. See eqn. (4) and (5).
was designed specifically for radical reactions and was based on
hydrogen abstraction from substituted cumenes by a polystyryl
20
radical. On the other hand, sigma plus was defined from the
solvolysis of X-cumyl chlorides. Nevertheless, its use in corre-
lating radical reactions has been shown to be widespread and
(
(
4)
5)
9
appropriate.
The strong dependence of cytotoxicity on ES-2 of the mostly
hydrophobic substituents on the phenyl ring, led to an examin-
ation of the role of hydrophobicity in this phenomenon. Des-
pite the collinearity between calculated partition coefficients
2
The addition of a steric parameter in QSAR 5 did not
improve the correlation; the lack of a substituent in the other
ortho (6-) position means that steric hindrance was not a factor
in the reactivity of the unencumbered 2-X-phenoxyl radical.
Continued interest in the radical forming ability of substi-
tuted phenols has led to a study of polysubstituted phenols and
in particular, analogs of the widely used antioxidants: butylated
hydroxyanisole (BHA) and butylated hydroxytoluene (BHT)
(Clog P) and ES-2, (r = 0.692), utilization of Clog P in lieu of
E
S-2 as in QSAR 6 gives rise to a much poorer correlation as
seen in QSAR 8.
(
7)
Hydrophobicity in tandem with ER leads to the following
QSAR:
(
Fig. 1).
(
8)
2
2
The significant differences in r (0.134) and q (0.266)
for QSAR 6 and 8 clearly establish the significance of steric
attributes in the generation of cytotoxicity.
Fig. 1 Structures of BHA, BHT and analogs
21,22
Recently, Garg et al.
have examined the various types of
BHA and BHT were generally recognized as safe additives in
toxicity wielded by X-phenols and sequestered them into three
main groups: toxicities governed by hydrophobicity alone, by
hydrophobicity and sigma minus parameters and by hydro-
phobicity and sigma plus parameters. Examination of this
data using these approaches led to the formulation of QSAR 9
and 10 which again, in terms of statistical significance, pale by
comparison to QSAR 6.
food by the FDA although these antioxidants have been impli-
cated in a number of toxic and carcinogenic events in animals.
BHT causes damage in alveolar and pulmonary endothelial
12
cells in the murine lung. It also induces hemorrhagic death
and liver necrosis in rats as well as carcinogenesis in mice and
13–15
rats.
The carcinogenicity of BHA has been well established
in rodents and other experimental animals but there is no data
16–19
on its carcinogenicity in man.
Although the toxic and
carcinogenic effects of BHT and BHA have been well docu-
mented, their mechanisms of action at the molecular level have
not been clearly elucidated, in that the role of various substitu-
ents on the phenolic ring have not been delineated. This study
examines the cytotoxic effects of a series of 4-X-2-alkyl or
(9)
4
-X-4,6-dialkylphenols versus L1210 leukemia cells in terms of
(
10)
their physico-chemical attributes.
Results
From a study of these hydrophobicity and Hammett sigma
based QSAR 7–10, it is clear that these particular ortho alkyl
substituted phenols behave in a fashion that deviates from most
other phenols.
The cytotoxicity data in Table 1 have led to the formulation of
QSAR 6.
The three outliers 2,6-(OCH ) -phenol, 2-CH -4-NO -phenol
3
2
3
2
and 2-CH -4-acetylphenol are mispredicted by a factor greater
3
(
6)
than four. The deviations of these three analogs are more than
three times the standard deviation of the equation (QSAR 6).
The 2,6-(OCH ) phenol is more active than predicted and is
3
2
also the only substituted phenol with alkoxy groups in both
ES-2 is Taft’s steric parameter for the larger of the two ortho
substituents. Since all substituent values of ES-2 are negative, the
negative coefficient with the ES-2 term in QSAR 6 indicates that
cytotoxicity is enhanced by larger substituents. Summation
ortho positions. This could be related to the conformations of
11
these phenols. Ingold et al. have shown via computational and
experimental data that steric repulsion exerted by ortho methyl
or ortho tert-butyl groups is not sufficient to force the OH bond
J. Chem. Soc., Perkin Trans. 2, 2002, 1112–1117
1113

View Article Online
Table 1 Cytotoxicity data and physico-chemical parameters of 2-alkyl and 2,6-dialkyl-4-X-phenols
log 1/C
b
Compound
Relative BDE /
kcal mol
a
ϩ
Ϫ1
c
number
Substituent
Observed
Predicted
Deviation
ER
ES-2
σ
Clog P
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2,6-di-Me
2,6-di-OMe
2,4,6-tri-Me
2,6-di-CMe3
2,6-di-CMe -4-Me
3.02
3.86
3.20
3.85
4.04
3.26
3.25
3.90
3.73
4.90
3.91
4.11
4.24
3.80
3.04
3.09
3.49
3.46
3.39
3.14
3.80
3.16
3.22
3.23
3.88
3.95
3.19
3.38
3.95
3.88
4.87
3.95
4.17
3.88
3.88
3.16
2.92
4.08
3.37
3.35
3.66
3.88
Ϫ0.14
0.64
Ϫ0.03
Ϫ0.03
0.09
0.06
Ϫ1.24
Ϫ0.55
Ϫ1.24
Ϫ2.78
Ϫ2.78
Ϫ1.31
Ϫ1.71
Ϫ2.78
Ϫ2.78
Ϫ2.78
Ϫ2.78
Ϫ2.78
Ϫ2.78
Ϫ2.78
Ϫ1.24
Ϫ1.24
Ϫ1.24
Ϫ1.24
Ϫ1.24
Ϫ1.24
Ϫ2.78
Ϫ0.62
Ϫ1.56
Ϫ0.93
Ϫ0.52
Ϫ0.83
Ϫ0.60
Ϫ0.56
Ϫ0.78
Ϫ0.57
0.27
Ϫ0.82
Ϫ0.37
Ϫ0.52
Ϫ0.57
Ϫ0.62
Ϫ0.38
0.48
Ϫ4.30
Ϫ5.80
Ϫ5.90
Ϫ8.00
Ϫ9.70
Ϫ4.90
Ϫ5.80
Ϫ9.30
Ϫ5.40
Ϫ3.70
Ϫ9.50
Ϫ7.90
Ϫ4.20
Ϫ4.50
Ϫ3.50
Ϫ3.70
2.40
2.37
1.11
2.87
4.93
5.43
3.43
3.93
6.75
3.65
5.31
5.96
6.09
5.03
3.70
2.42
2.36
2.30
3.08
2.02
1.90
4.23
d
0.22
0.09
0.06
0.09
0.06
0.06
0.09
0.06
0.47
0.09
0.18
0.06
0.06
0.06
Ϫ0.04
0.44
0.15
0.14
0.27
0.06
3
2,6-di-C H
0.07
2
5
2,6-di-CHMe2
2,4,6-tri-CMe3
Ϫ0.13
Ϫ0.05
Ϫ0.15
0.03
Ϫ0.04
Ϫ0.06
0.37
Ϫ0.08
Ϫ0.12
0.18
Ϫ0.57
0.09
2-CMe -6-Me
3
1
1
1
1
1
1
1
1
1
1
2
2
2,6-di-CMe -4-NO
2,6-di-CMe -4-C H
2,6-di-CMe -4-Br
2,4-di-CMe3
3
2
3
2
5
3
2-CMe -4-Me
3
2,4-di-Me
2-Me-4-F
2-Me-4-NO2
2-Me-4-Br
2-Me-4-OMe
d
Ϫ0.16
Ϫ1.09
0.33
Ϫ2.10
Ϫ6.90
—
0.04
Ϫ0.52
Ϫ0.08
d
2-Me-4-COMe
2-CMe -4-C H
0.16
Ϫ4.40
3
2
5
a
b
c
d
Predicted using QSAR 6. Relative to phenol, BDE = 0. Calculated using Clog P Version 4.0. Data points omitted in formulation of QSAR 6
out of the plane of the aromatic ring. Thus the phenolic OH
hydrogen can “tuck” itself between two out of plane H or CH3
Photooxidation quenching of singlet oxygen from 4-X-2,6-di-
25
tert-butylphenols in methanol.
groups while the third H or CH group can lie in the plane of
3
the aromatic ring. This manoeuvring destabilizes the phenol
relative to the phenoxyl radical. In the case of the BHA/BHT
analogs, compounds 1 and 3–21 would be subject to similar
changes in geometry.
(
14)
These examples illustrate that for a variety of radical reac-
tions under differing conditions, σ gives excellent correlations
Compound 2 is an anomaly in terms of its conformational
constraints and its biological activity. o-Methoxyphenols have
two conformations—one with the methoxy group facing the
phenolic group and one with it in the away position. The former
position is stabilized by internal hydrogen bonding. For o,o-
dimethoxyphenols in which internal hydrogen bonding must
occur, one methoxy group would point away from and one
towards the OH moiety. However, in the corresponding radical,
both groups would point away from the phenoxyl radical’s oxy-
gen atom leading to its destabilization and perhaps its enhanced
reactivity. The aberrant behavior of compounds 17 (2-Me-4-
NO ) and 20 (2-Me-4-COCH ) cannot be explained; the nitro
ϩ
with negative ρ values. Note the lack of a steric parameter in
QSARs 11–14. This is attributed to the constancy of the steric
effect in the 2 and 6 positions where all the congeners bear
bulky tert-butyl groups in these positions. Variation in the steric
effect at the para position has no impact on phenoxyl radical
ϩ
formation. These ρ values are contradictory to those observed
with the BHA and BHT analogs. It is apparent that the
most bulky multi-substituted phenols of QSAR 6 are behaving
in a manner distinct from those supporting QSAR 1 or QSAR
1
1–14. These results suggest that the use of ER is more
2
3
appropriate in a sterically demanding environment.
and acetyl groups may be subject to enhanced metabolism in
the cells.
Further insight on the reactions of radicals with 2,6-di-tert-
butylphenols with substituents in the 4-position is revealed by
the following examples: QSARs 11–14.
Discussion
In QSAR 6 there are two nitro containing congeners one of
which is well fitted, while the other having only one small ortho
substituent (2-Me, 4-NO ) is poorly fitted. This latter com-
Hydrogen abstraction from 4-X-2,6-di-tert-butylphenols by 5-
2
23
butylperoxyl radicals in isopentane at Ϫ37 ЊC.
pound behaves more like those of QSAR 1 where the cyto-
toxicities of simple phenols having only one ortho substituent
are not well predicted unless log P is set to zero (QSAR 4). An
attempt to combine the simple mono-substituted ortho phenols
of QSAR 4 with the polysubstituted phenols in QSAR 6
resulted in a poor correlation. In QSAR 6, ES-2 is the most
significant parameter accounting for 70% of the variance while
ER accounts for 23%. There are other radical parameters, but
(
11)
Hydrogen abstraction from 4-X-2,6-di-tert-butylphenols by
24
styrene peroxide radicals at 65 ЊC.
26
ؒ
none that cover as wide a range as E . Use of σ defined by
R
(
12)
27
Cleary et al. reduces the data set to 14 data points because of
ؒ
ؒ
the paucity of established σ values. Use of σ in lieu of E
R
2
yields r of 0.902, with ρ = ϩ1.54, which is essentially the same
Hydrogen abstraction from 4-X-2,6-di-tert-butylphenols by
tetralin peroxide (1,2,3,4-tetrahydro-1-naphthyl hydroperoxide)
quality of fit as E . The strong dependence on Otsu’s E and
R
R
ؒ
24
Cleary’s σ suggests that these reactions may be radical medi-
ated. The ortho-substituted complex phenols in the current
study behave quite differently from the phenols that have been
studied previously (QSAR 1 and 2), and even from the small set
of ortho-substituted electron-rich phenols that were examined
radicals at 65 ЊC.
(
13)
1
114
J. Chem. Soc., Perkin Trans. 2, 2002, 1112–1117

View Article Online
ϩ
later (QSAR 4 and 5). This difference in behavior is difficult to
explain, but it is not unexpected. Radical-mediated phenolic
toxicity is a multi-step process and large structural variations
may alter the nature of the rate-determining step and may even
lead to a change in mechanism. It is possible that different
biological radicals are responsible for activation of sterically
hindered phenols, and the resulting hindered phenoxyl radicals
may interact with a different set of cellular targets than their
less hindered congeners.
The ambiguity of the different roles of σ and ER in the
correlation of radicals warrants further extensive investigation.
A number of instances where ER is a better measure of sub-
stituent effects have been identified (ref. 9, Table 7) but no clear
guidelines emerge to establish what structural features select for
one parameter over the other. The current study suggests that
transition state geometry, especially steric hindrance, may be
important. Crowding around the phenolic OH affects the ease
ؒ
of H removal, the lifetime of the phenoxyl radical, and the ease
An example of how ortho substituents affect the magnitude,
if not the type, of substituent effect is given by phenol ioniz-
ation (QSAR 15–17).
of phenoxyl attack on cellular targets, and any one or more of
these factors may alter the nature of substituent effects.
The properties of the OH bond appear to be playing a signifi-
cant role in the reactivities of phenolic compounds and their
subsequent cytotoxicities. Despite the large number of newer
experimental techniques and the design of computational tools
that have addressed the kinetic behavior and thermodynamic
stabilities of mono, di, tri and multisubstituted phenols, there
are still derivatives that defy characterization such as 2,6-
Ionization of 4-X-2,6-di-tert-butyl phenols in 50 aqueous
28
methanol.
35
(
15)
dimethoxyphenol. In our cytotoxicity study, this phenol is
also an outlier. Perhaps its electronic/steric attributes have not
been adequately delineated.
Attempts to derive meaningful QSAR of multisubstituted
phenols with BDE or σ yielded poor results even after remov-
29
Ionization of simple X-phenols in 51 aqueous ethanol.
ϩ
ing all but five of nine such compounds. The fact that
2
,4-disubstituted phenols fitted QSAR 6 while 2-substituted
phenols did not, suggests that E is better suited for sterically
(
16)
17)
R
hindered 2,4,6-trisubstituted phenols. Clearly the toxicity of
phenols is a multifaceted problem where the dependent variable
log 1/C may not be uniform. Until this can be established one
cannot gather all phenols together to obtain a single meaning-
ful equation. However, QSAR 1 and 6 provide considerable
insight into a complex problem and will provide valuable
guidance for future studies.
30
Ionization of simple phenols in aqueous solution.
(
Experimental
Melting points were determined on an electrothermal melting
The intercepts of these equations show that 2,6-disubstituted
phenols are considerably less acidic [eqn. (15)] than less hin-
dered phenols [eqn. (16), (17)] even after solvent effects on pKa
are taken into account. The loss of acidity can be attributed to
steric hindrance causing loss of solvation (loss of hydrogen
bonding) in the 2,6-disubstituted phenoxide anions. The same
factor also explains the enhanced effect that substituents have
on the ionization of 2,6-disubstituted phenols [compare slopes
for eqn. (15)–(17)]. Loss of hydrogen bonding promotes charge
delocalization in the phenoxide anion, making this anion more
susceptible to substituent effects. QSAR 15–17 highlight the
difficulty of proton loss in ionization. A similar problem may
1
point apparatus (MEL–TEMP II with digital thermometer). H
13
NMR and C NMR spectra were recorded on a Bruker DPX
400 MHz NMR spectrometer with TMS as the internal stand-
ard; chemical shifts are given in δ (ppm) and coupling con-
stants, J, in MHz. Infrared (IR) spectra were recorded on a
Perkin Elmer 1600 series FTIR and only principal, sharply
defined IR peaks are reported. Mass spectra were obtained
from GC/MS data from a Hewlett Packard GC/MS system HP
6890 series with a mass selective detector. Thin-layer chrom-
atography was performed on silica gel plates (silica gel IB-F
Baker). Data for chemical elemental analysis were obtained
from Desert Analytics (Tucson, AZ).
ؒ
occur in H abstraction.
It has been well established by Bradbury and Lipnick that a
large number of simple organic molecules may be classified as
3
1
narcotics. Narcosis may be classified into two types: Type 1,
Materials
32
nonpolar narcosis and Type 2 designated as polar narcosis.
Most of the phenols are commercially available compounds
except: 2,6-di-tert-butylhydroquinone (2,6-di-tert-butyl-4-
hydroxyphenol), 2,6-di-tert-butyl-4-acetoxyphenol; 2,6-di-tert-
Nonspecific toxicity which is generally exhibited by nonpolar
narcotics is directly proportional to the concentration at the site
of action which is usually the membrane. This toxicity can be
directly predicted from their hydrophobicities by QSAR. This is
seen in QSAR 7. However, in this case the toxicity is greater
than this baseline approach since only 59% of the variance in
the data can be explained by hydrophobicity. Thus it is clear
that the phenols in this case are not acting via a Type 1 narcosis.
Extensive studies of phenols acting via a Type 2 polar narcosis
mechanism of action have been carried out by Cronin and
Schultz who developed a two-parameter approach (log P and
butyl-4-nitrophenol;
2,6-diethylphenol;
2-methyl-4-nitro-
phenol; 4-methoxy-2-methylphenol and 4-bromo-2-methyl-
phenol, which were synthesized as follows.
Syntheses
2,6-Di-tert-butylhydroquinone. 2,6-Di-tert-butylhydroquin-
one was synthesized by the oxidation of 4-hydroxy-3,5-di-tert-
butylbenzaldehyde with hydrogen peroxide. Commercially
available 4-hydroxy-3,5-di-tert-butylbenzaldehyde (1.93g; 8.25
mmol) was dissolved in pyridine (7 ml) and treated with aque-
ous potassium hydroxide (500 mg of KOH dissolved in l ml of
33,34
ELUMO) to deal with the “excess” toxicity of phenols.
This
model provides a mechanistic basis for toxicity in terms of par-
titioning and bioreactivity. A similar approach to the results in
Table 1 resulted in the formulation of QSAR 9, which only
explains 80% of the variance in the data. This suggests that
other interactions are at play as indicated by QSAR 6 where
steric and radical reactivity parameters are significant.
H O) under nitrogen. Hydrogen peroxide (1.5 ml, 30% v/v) was
2
added dropwise to the reaction mixture over a period of 30 min
with constant stirring under nitrogen. After the addition of
H O , the reaction mixture was refluxed with constant stirring
2
2
J. Chem. Soc., Perkin Trans. 2, 2002, 1112–1117
1115

View Article Online
Ϫ1
for 40 min under nitrogen and then cooled to room temper-
ature. The reaction mixture was poured into 6 M HCl (18 ml)
and extracted with ether. The ether layer was washed with water
125.794 (C3), 126.64 (C2), 139.576 (C4), 162.52 (C1); νmax/cm
ϩ ϩ
3461.4 (OH); m/z 154 (M ϩ 1, 9.5%), 153 (M , 100), 123 (35),
77 (84).
twice and dried over MgSO (anhyd.). The solvent was removed
4
and the crude product was crystallized from petroleum ether
4
-Methoxy-2-methylphenol. 4-Methoxy-2-methylphenol was
(
bp 35–60 ЊC) After three consecutive crystallizations, the pure
obtained from the diazotization of 4-methoxy-2-methyl-
aniline. The crude product was crystallized from a 1 : 1
product of 2,6-di-tert-butylhydroquinone (1.1 g; 60.11%) was
42
36–38
obtained, mp 104–105 ЊC;
δH (400 MHz; DMSO-d ) 1.35
6
mixture of CH Cl –petroleum ether in 58% yield; mp 64–65
2
2
(
18H, s, 6 CH ), 6.3 (1H, s, OH), 6.5 (2H, s, Ph), 8.55 (1H, s,
44
3
ЊC; δH (400 MHz; DMSO-d ) 2.0 (s, 3H, CH ), 3.55 (s, 3H,
6
3
OH); δ (400 MHz, DMSO-d ) 30.775 (6CH ), 34.934 (2-C<),
C
6
3
OCH ), 6.47 (d, J = 2.8, 1H, Ph), 6.58 (d, J = 8.5, 2H, Ph), 8.67
3
s, 1H, OH); δC (400 MHz; DMSO-d ) 17.155 (CH ), 54.817
(OCH ), 112.281 (C5), 114.772 (C3), 115.941 (C6), 124.971
1
11.362 (C3 & C5), 141.558 (C2 & C6), 146.344 (C1), 150.726
Ϫ1 ϩ
(
6
3
(
C4); νmax/cm 3343.5 (OH), 3632.8 (OH); m/z 223 (M ϩ 1,
3
ϩ
6
%), 222 (M , 39), 207 (100), 82 (21), 57 (68) (Calc. C H O ;
Ϫ1
14
22
2
(C2), 149.446 (C1), 152.234 (C4); νmax/cm 3280 (OH); m/z 139
C, 75.68; H, 9.91. Found: C, 75.39; H, 9.88%).
ϩ
M
ϩ
(
ϩ 1, 9.5%), 138 (M , 90.5), 123 (100), 95 (15), 77 (19),
6
7 (24).
2
,6-Di-tert-butyl-4-acetoxyphenol. Commercially available
2
,6-di-tert-butyl-p-benzoquinone (550 mg, 2.5 mmol) was dis-
4
-Bromo-2-methylphenol. 4-Bromo-2-methylphenol was also
solved in acetic anhydride (3 ml) and Zn dust (150 mg) was
added followed by pyridine (0.5 ml). The reaction mixture was
stirred at room temperature for 3 hours and filtered. To the
filtrate, water (5 ml) was added and stirred for 30 minutes. The
crude product precipitated, and was washed with water and
dried. It was crystallized from petroleum ether (bp 35–60Њ) to
give 2,6-di-tert-butyl-4-acetoxyphenol (400 mg, 60.5%), mp
synthesized by a diazonium reaction of 4-bromo-2-methyl-
aniline and the crude product was crystallized from petroleum
ether in 32.1% yield. Mp 61 ЊC; δH (400 MHz; DMSO-d6)
.09 (s, 3H, CH ), 6.72 (d, J = 8.4, 1H, Ph), 7.13 (d, J = 8.4, 1H,
Ph), 7.22 (s, 1H, Ph), 9.62 (s, 1H, OH); δ (400 MHz; DMSO-
d ) 16.942 (CH ), 110.93 (C4), 117.758 (C6), 128.136 (C2),
30.387 (C5), 133.993 (C3), 156.057 (C1); νmax/cm 3260.8
OH); m/z 188 (M ϩ 1, 81.8%), 187 (M , 15.2), 186 (87.9), 107
100), 77 (72.8).
42
44
2
3
C
6
3
Ϫ1
1
(
(
39,40
9
0–95.5 ЊC;
δH (400 MHz; DMSO-d ) 1.4 (18H, s, 6CH ),
6
3
ϩ
ϩ
2
.25 (3H, s, OCOCH ), 6.82 (2H, s, Ph), 7.05 (IH, s, OH);
3
δ (400 MHz; DMSO-d ) 19.195 (CH ), 28.465 (6CH ), 32.954
C
6
3
3
(
(
2-C<), 115.799 (C3 & C5), 139.084 (C2 & C6), 141.865
Ϫ1
Cytotoxicity evaluation
C4), 149.626 (Cl), 167.967 (CO); νmax/cm 3587.1 (OH);
ϩ
ϩ
m/z 265 (M ϩ 1, 3%), 264 (M , 12), 222 (94), 207 (100), 57
L1210 cells were maintained in asynchronous logarithmic
growth at 37 ЊC in RPMI medium with -glutamine sup-
plemented with 10% (v/v) FBS. All stock solutions and
dilutions were made in unsupplemented RPMI medium.
(
54) (Calc. C H O ; C, 72.73; H, 9.09. Found: C, 72.75; H,
16 24 3
9
.07%).
2
,6-Di-tert-butyl-4-nitrophenol.
phenol was synthesized by the nitration of 2,6-di-tert-butyl-
phenol with CH COOH–HNO3 in cyclohexane. A mixture
2,6-Di-tert-butyl-4-nitro-
4
Ϫ1
Cell cultures were seeded at 2–5 × 10 cells ml in duplicate
for each inhibitor concentration in a 96 well microtiter plate
(180 µl per well). The test compounds (20 µl) were then added to
the cell cultures in 1 : 10 dilution in order to achieve the desired
concentration. Each inhibitor was tested at a minimum of 8
concentrations. After 48 hours of continuous drug exposure,
the cells were counted by using the CyQUANT GR assay kit
from Molecular Probes. For this purpose, the medium was
removed from the plates which were then frozen at Ϫ80 ЊC for a
minimum of 1 h. The cells were thawed at 37 ЊC and 200 µl of
CyQUANT GR dye/cell lysis buffer was added to each well.
The plates were incubated for 5 minutes at 37 ЊC and their
fluorescence was measured using a Cytofluor II multiwell
fluorescence plate reader. The excitation maximum was 485 nm
and the emission maximum was 530 nm. From the data, a dose
response curve was drawn and the IC50 determined. The
CyQUANT GR assay measures the ability of CyQUANT GR
dye to bind to the cellular nucleic acids of viable cells. Cyto-
toxicity is expressed as the concentration of the phenol (IC )
3
of CH COOH–HNO (3 ml; 1 : 1 v/v) was added dropwise
3
3
to a stirred solution of 2,6-di-tert-butylphenol (5.15 g; 25
mmol) in cyclohexane (15 ml) by maintaining the temperature
at 10–20 ЊC. The reaction mixture was further stirred at this
temperature for 10 min and the precipitate obtained was fil-
tered, washed with water and dried. The crude product was
crystallized from petroleum ether and gave the purified product
41
as white needles (3.74 g; 59.6%); mp 155–156 ЊC; δH (400
MHz; DMSO-d ) 1.33 (18H, s, 6CH ), 7.9 (2H, s, Ph), 8.45
6
3
(
1H, s, OH); δ (400 MHz; DMSO-d ) 29.911 (6CH ), 35.033
C
6
3
(
2-C<), 120.907 (C3 & C5), 139.384 (C2 & C6), 142.82 (C4),
Ϫ
1
ϩ
1
2
57.53 (Cl); νmax/cm 3626.5 (OH); m/z 252 (M ϩ 1, 3.08%),
ϩ
51 (M , 17), 237 (14), 236 (100), 208 (29.3) (Calc. C H NO ;
14
21
3
C, 66.93; H, 8.36; N, 5.57. Found: C, 67.17; H, 8.21; N,
.66%).
5
50
2
,6-Diethylphenol. 2,6-Diethylphenol was synthesized from
42
that causes a 50% reduction in fluorescence as compared with
commercially available 2,6-diethylaniline by a diazo reaction.
The product was obtained in 45.3% yield, mp 38 ЊC; δH (400
MHz; DMSO-d ) 1.115 (t, J = 7.75, 6H, 2CH ), 2.574 (q, 4H,
46
the controls. Physicochemical constants were then utilized to
43
formulate
a
quantitative structure–activity relationship
6
3
(QSAR) for these compounds.
2
CH ), 6.70 (t, J = 7.73, 1H, Ph), 6.95 (d, J = 6.97, 2H, Ph), 8.07
2
(
s, 1H, OH); δ (400 MHz; DMSO-d ) 14.773 (2CH ), 23.284
C
6
3
Computational methods
(
2CH ), 119.843 (C4), 126.741 (C3 & C5), 130.76 (C2 & C6),
2
Ϫ1 ϩ
1
1
52.373 (C1); νmax/cm 3362 (OH); m/z 151 (M ϩ 1, 4.6%),
BDE was defined as the energy of the following reaction: X–
ϩ
ؒ
ؒ
50 (M , 41.58), 135 (100), 121 (35.44), 91 (33.9); (Calc.
PhOH ϩ PhO
X–PhO ϩ PhOH using RB3LYP/6-31G**
C H O; C, 80; H, 9.33. Found: C, 79.43; H, 9.27%).
energies for the phenols and UB3LYP/6-31G** energies for the
10
14
radicals (obtained using Jaguar 4.1, “accurate” pseudospectral
Ϫ1 47,48
2
-Methyl-4-nitrophenol. 2-Methyl-4-nitroaniline was con-
grids, numerical errors <0.2 kcal mol ).
Initial phenol
42
verted into 2-methyl-4-nitrophenol by the diazonium method.
The crude product was crystallized from a 1 : 1 mixture of
CCl –petroleum ether in 65.4% yield; mp 94 ЊC;
MHz; DMSO-d ) 2.1 (s, 3H, CH ), 6.9 (d, J = 6.7, 1H, Ph), 7.92
d, J = 6.7, 1H, Ph), 8.1 (s, 1H, Ph), 11.0 (br s, 1H, OH); δ (400
C
MHz; DMSO-d ) 16.068 (CH ), 114.915 (C6), 124.042 (C5),
geometries were obtained from an automatic conformer search
4
8
4
9
(MMFF force field using PC Spartan Pro, v. 1.0.5) followed
by RB3LYP/6-31G** geometry optimization of each con-
former. The lowest energy conformer (after RB3LYP optimiz-
ation) was used to calculate BDE and to generate the initial
radical geometry.
44,45
δH (400
4
6
3
(
6
3
1
116
J. Chem. Soc., Perkin Trans. 2, 2002, 1112–1117

View Article Online
2
2
4 J. A. Howard and K. U. Ingold, Can. J. Chem., 1965, 43, 2724.
5 M. J. Thomas and C. S. Foote, Photochem. Photobiol., 1978, 27, 683.
Acknowledgements
This research was supported by NIEHS grant R01 ES 07595
and in part by NSF-CHE-9724418.
26 C. Hansch and A. Leo, Exploring QSAR. Fundamentals and
Applications in Chemistry and Biology, American Chemical Society,
Washington DC, 1995.
2
7 X. Cleary, M. E. Mehrsheikh-Mohammad and S. McDonald,
J. Org. Chem., 1987, 52, 3254.
References
2
8 G. Schwarzenbach and H. Egli, Helv. Chim. Acta, 1934, 17, 1176.
1
2
H. F. Stich, Mutat. Res., 1991, 259, 307.
29 L. A. Cohen and W. M. Jones, J. Am. Chem. Soc., 1963, 85, 3397.
30 J. W. Bunting and D. Stefanidis, J. Am. Chem. Soc., 1988, 110, 4008.
31 S. P. Bradbury and R. L. Lipnick, Environ. Health Perspect., 1990,
87, 181.
32 G. D. Veith and S. J. Broderius, Environ. Health Perspect., 1990, 87,
207.
A. Arora, M. G. Nair and G. M. Strasburg, Free Radical Biol. Med.,
1
998, 24, 1355.
3
T. Ohnishi, H. Yamazaki, T. Iyanagi, T. Nakamura and I. Yamazaki,
Biochim. Biophys. Acta, 1969, 172, 357.
4
5
P. H. Jellinck and R. Fletcher, Can. J. Biochem., 1971, 49, 885.
C. D. Selassie, A. J. Shusterman, S. Kapur, R. P. Verma, L. Zhang
and C. Hansch, J. Chem. Soc., Perkin Trans. 2, 1999, 2729.
F. G. Bordwell, Z. M. Zhang, A. V. Satish and J. P. Chang, J. Am.
Chem. Soc., 1994, 116, 6605.
33 M. T. D. Cronin and T. W. Schultz, Chem. Res. Toxicol., 2001, 14,
1284.
6
7
8
9
34 M. T. D. Cronin and T. W. Schultz, Sci. Total Environ., 1997, 204, 75.
35 R. M. Borges dos Santos and J. A. M. Simoes, J. Phys. Chem. Ref.
Data, 1998, 27, 707.
36 G. A. Nikoforov and V. V. Ershov, Izv. Akad. Nauk SSSR, Ser
Khim., 1964, 176 (Chem. Abstr., 1964, 60, 9188g).
37 E. Muller and K. Ley, Chem. Ber., 1955, 88, 601 (Chem. Abstr., 1956,
50, 5558a).
T. Brinck, M. Haeberlein and M. Jonsson, J. Am. Chem. Soc., 1997,
1
19, 4239.
S. Tomiyama, S. Sakai, T. Nishiyama and F. Yamada, Bull. Chem.
Soc. Jpn., 1993, 66, 299.
C. Hansch and H. Gao, Chem. Rev., 1997, 97, 2995.
1
1
1
1
1
0 C. D. Selassie, T. V. DeSoyza, M. Rosario, H. Gao and C. Hansch,
38 D. H. R. Barton, P. G. Gordon and D. G. Hewitt, J. Chem. Soc. (C),
1971, 1206.
Chem.-Biol. Interact., 1998, 113, 175.
1 J. S. Wright, D. J. Carpenter, D. J. McKay and K. U. Ingold, J. Am.
Chem. Soc., 1997, 119, 4245.
2 A. A. Marino and J. T. Mitchell, Proc. Soc. Exp. Biol. Med., 1972,
40, 122.
3 O. Takahashi and K. Hiraga, Toxicol. Appl. Pharmacol., 1978, 43,
99.
4 Y. Nakagawa, K. Tayama, T. Nakao and K. Hiraga, Biochem.
Pharmacol., 1984, 33, 2667.
5 H. Witschi and S. Lock, Toxicol. Appl. Pharmacol., 1979, 50, 391.
6 H. Verhagen, Pharm. Weekbl., Sci. Ed., 1990, 12, 164.
7 E. A. Nera, F. Iverson, E. Lok, C. L. Armstrong, K. Karpinsky and
D. B. Clayson, Toxicology, 1988, 53, 251.
39 M. Lounasmaa, Acta. Chem. Scand., 1967, 21, 2807 (Chem. Abstr.,
1968, 69, 35634m).
40 Z. A. Dobronravova, V. I. Meshchevyakov, Yu. E. Prishchenko,
L. D. Gavrilov and L. I. Vereshchagin, Zh. Org. Khim., 1980, 16,
1034 (Chem. Abstr., 1980, 93, 167786t).
1
3
41 J. S. Meek, J. S. Fowler, P. A. Monroe and T. J. Clark, J. Org. Chem.,
1968, 33, 223.
42 B. S. Furniss, A. J. Hannaford, P. W. G. Smith and A. R. Tatchell,
Vogel’s Textbook of Practical Organic Chemistry, Fifth Edition,
Longman Scientific & Technical, 1989, p. 927.
43 G. E. George and J. K. Alfred, USP 3 075 832/1963 (Chem. Abstr.,
1963, 59, 12707a).
1
1
1
1
1
2
2
8 N. Ito, S. Fukushima, A. Hagiwara, M. Shibata and T. Ogiso,
J. Natl. Cancer Inst., 1983, 70, 343.
9 H. Verhagen, P. A. E. L. Schilderman and J. C. S. Kleinjans,
Chem.-Biol. Interact., 1991, 80, 109.
0 T. Otsu, T. Ito, Y. Fuji and M. Imoto, Bull. Chem. Soc. Jpn., 1968,
44 M. Fujio, M. Mishima, Y. Tsuno, Y. Yukawa and Y. Takai, Bull.
Chem. Soc. Jpn., 1975, 48, 2127.
45 M. J. Collins, P. M. Hatton and S. Sternhell, Aust. J. Chem., 1992,
45, 1119.
46 L. J. Jones, M. Gray, S. T. Yue, R. P. Haugland and V. L. Singer,
J. Immunol. Methods, 2001, 254, 85.
47 Jaguar 4.1, Schrodinger, Inc., Portland, OR, 2000.
48 T. A. Halgren, J. Comput. Chem., 1996, 17, 490.
49 PC Spartan Pro 1.0.5, Wavefunction, Inc., 18401 Von Karman Ave,
Suite 370, Irvine, CA 92612, USA.
4
1, 204.
1 R. Garg, A. Kurup and C. Hansch, Crit. Rev. Toxicol., 2001, 31,
23.
2
2
2 R. Garg, S. Kapur and C. Hansch, Med. Res. Rev., 2001, 21, 73.
2
3 J. A. Howard and E. Furimsky, Can. J. Chem., 1973, 51, 3738.
J. Chem. Soc., Perkin Trans. 2, 2002, 1112–1117
1117