Cellular apoptosis and cytotoxicity of phenolic compounds: A quantitative structure-activity relationship study

Article abstract of DOI:10.1021/jm050567w

In this comprehensive study on the caspase-mediated apoptosis-inducing effect of 51 substituted phenols in a murine leukemia cell line (L1210), we determined the concentrations needed to induce caspase activity by 50% (I ₅₀) and utilized these

Full text of DOI:10.1021/jm050567w

7
234
J. Med. Chem. 2005, 48, 7234-7242
Cellular Apoptosis and Cytotoxicity of Phenolic Compounds: A Quantitative
Structure-Activity Relationship Study
Cynthia D. Selassie,* Sanjay Kapur, Rajeshwar P. Verma, and Melissa Rosario
Department of Chemistry, Pomona College, 645 North College Avenue, Claremont, California 91711
Received June 16, 2005
In this comprehensive study on the caspase-mediated apoptosis-inducing effect of 51 substituted
phenols in a murine leukemia cell line (L1210), we determined the concentrations needed to
induce caspase activity by 50% (I50) and utilized these data to develop the following quantitative
1
structure-activity relationship (QSAR) model: log /I50 ) 1.06 B5
2
+ 0.33 B5 - 0.18π2,4 - 0.92.
3
3 2
B5 and B5 represent steric terms, while π2,4 represents the hydrophobic character of the
substituents on the ring. The strong dependence of caspase-mediated apoptosis on mostly steric
parameters suggests that the process is a receptor-mediated interaction with caspases or
mitochondrial proteins being the likely targets. Conversely, cytotoxicity studies of 65 electron-
1
releasing phenols in the L1210 cell line led to the development of the following equation: log /
+
ID50 ) -1.39σ - 0.28 B52,6 + 0.16 log P - 0.58I
2
- 1.04I + 3.90. The low coefficient with log P
1
may pertain to cellular transport that may be enhanced by a modest increase in overall
+
hydrophobicity, while the presence of σ is consistent with the suggestion that radical
stabilization is of prime importance in the case of electron-releasing substituents. On the other
1
hand, the QSAR for the interactions of 27 electron-attracting phenols in L1210 cells, log /ID50
)
0.56 log P - 0.30 B5 + 2.79, suggests that hydrophobicity, as represented by log P is of
2
critical importance. Similar cytotoxicity patterns are observed in other mammalian cell lines
such as HL-60, MCF-7, CCRF-CEM, and CEM/VLB. The significant differences between the
cytotoxicity and apoptosis QSAR for electron-releasing phenols suggest that cytotoxicity involves
minimal apoptosis in most of these substituted monophenols.
Introduction
cals as chemopreventive agents coupled with the ex-
tensive usage and prevalence of various phenols as
industrial chemicals or environmental pollutants war-
rants a thorough and systematic study of the variables
that define their chemical reactivities and subsequent
biological activities.10
In recent years, rapid advances have been made in
delineating the molecular mechanisms underlying apo-
ptosis or programmed cell death.1,2 Apoptosis is mostly
mediated through the activation of caspases that are
aspartate-specific cysteine proteases that target a wide
spectrum of specific proteins for limited proteolysis.3
Two major pathways of caspase activation have been
identified: a transmembrane receptor-mediated extrin-
sic pathway initiated by the binding of death activator
proteins and an intrinsic pathway targeting mitochon-
drial membranes that results in the release of factors
such as cytochrome c that promote caspase-9 activation.5
It is well established that perturbations in cell cycle or
metabolism, oxidative stress, growth factor availability,
and genotoxic agents that diminish mitochondrial in-
,4
In recent years, work in our laboratory has focused
on the myriad biological activities of X-phenols and their
ability to form radical species.11 Brown’s variant (σ ) of
Hammett’s electronic σ constant has been particularly
pertinent in delineating quantitative structure-activity
relationships (QSARs) pertaining to the cytotoxicity of
various substituted, albeit homogeneous, phenols in
murine leukemia cells. These early QSAR models are
+
12
described in eqs 1 and 2. Bond dissociation energies
as represented by BDE speak directly of the homolytic
cleavage of the O-H bond and subsequent formation
of the phenoxyl radical, a reactive oxygen species. In
both cases, the low coefficient with the hydrophobic term
6
tegrity, can also initiate cell death. Many chemothera-
peutic agents activate apoptosis, an activity often
mediated by the p53 pathway.
7
-9
The ability of various phenolic compounds to play
conflicting and complex roles as radical scavengers,
antioxidants, and prooxidants provides a rationale for
an examination of their apoptotic activities. Phenolic
compounds are ubiquitous in nature. They are univer-
sally distributed throughout the plant kingdom as
mostly hydroxylated aromatic rings that may also be
conjugated with other natural products such as fatty
acids and flavanoids or polymerized into larger entities
such as lignins. The use of phenolic-based phytochemi-
(
log P) suggests that the radical may be interacting with
a receptor such as DNA or it may represent the slightly
enhanced transport of the X-phenoxy radical in the
cellular environs. Phenols with substituents of an
electron-withdrawing nature do not subscribe to this
type of toxicity. Their mostly nonspecific cytotoxicities
are modeled by hydrophobicity, as reflected in eq 3.
In these QSAR equations, ID50 represents the molar
concentration of X-phenol that induces 50% growth
13
inhibition versus murine leukemia L1210 cells, and
thus, log /ID is the dependent variable that defines the
biological activity in QSAR studies. σ represents
1
*
To whom correspondence should be addressed. Phone: (909) 621-
50
+
8
446. Fax: (909) 607-7726. E-mail: cselassie@pomona.edu.
1
0.1021/jm050567w CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/14/2005

QSAR Study of Phenolic Compounds
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7235
Cytotoxicity of electron-releasing X-phenols
Table 1. Caspase-Mediated Apoptosis of Phenols Against
L1210 Cells
in L1210 cells:
log 1/ID50
1
+
log /
) (-1.35σ ( 0.15) + (0.18 ( 0.04)log P +
ID50
obsd preda B5
no.
X
2
π
2,4
B5
3
(
3.31 ( 0.11) (1)
1
2
3
4
5
6
7
8
9
4-OCH
4-CN
3
0.02 0.48 1.00 -0.02 1.00
0.10 0.58 1.00 -0.57 1.00
1.00 0.53 1.00 -0.28 1.00
2
2
n ) 51, r ) 0.895, s ) 0.227, q ) 0.882
4-NO
2
4-OC
4-OC
4-OC
4
H
H
H
9
5
7
0.14 0.20 1.00
0.19 0.10 1.00
0.09 0.29 1.00
0.11 0.12 1.00
1.52 1.00
2.08 1.00
1.05 1.00
1.98 1.00
1
log /
) (-0.19 ( 0.02)BDE +
6
3
ID50
(
2
0.21 ( 0.03)log P + (3.11 ( 0.10) (2)
3 3
4-C(CH )
4-COCH
H
3
0.03 0.58 1.00 -0.55 1.00
2
n ) 52, r ) 0.920, s ) 0.202, q ) 0.909
-0.20 0.48 1.00
0.79 0.96 1.00
1.10 1.35 1.00
0.74 0.74 1.00
1.00 0.79 1.00
0.82 0.59 1.00
1.25 0.80 1.00
1.11 0.68 1.00
1.31 1.17 1.00
0.45 0.82 1.00
0.79 0.79 1.00
2.80 2.53 3.17
2.43 2.23 3.17
2.49 2.54 3.17
1.90 2.42 3.17
2.90 2.46 3.17
2.50 2.54 3.17
2.10 2.05 3.17
2.88 2.52 3.17
2.39 2.61 3.17
2.41 2.52 3.17
2.58 2.26 3.17
2.09 2.31 3.17
2.49 2.47 3.17
2.87 2.42 3.17
0
0
0
0
0
0
0
0
0
0
0
1.00
2.44
3.61
1.80
1.95
1.35
1.97
1.60
3.07
2.04
1.93
1
0
3-NO
2
Cytotoxicity of electron-attracting X-phenols
11 3-NHCOCH
3
1
1
2
3
3-Cl
3-Br
in L1210 cells:
1
14 3-F
15 3-NH
2
log /
) (0.62 ( 0.16)log P + (2.35 ( 0.31) (3)
ID50
1
1
1
1
2
6
7
8
9
0
3-CN
3-OCH
3-CH
3-OH
2,6-(C(CH
2
2
n ) 15, r ) 0.845, s ) 0.232, q ) 0.800
3
3
Brown’s refinement of the Hammett σ constant, while
log P is a calculated partition coefficient (octanol-water
system).
3
)
3
)
2
,4-OCOCH
3
1.34 1.00
3.00 1.00
1.33 1.00
1.98 1.00
1.77 1.00
1.31 1.00
3.96 1.00
1.41 1.00
0.95 1.00
1.43 1.00
2.84 1.00
2.54 1.00
1.70 1.00
1.96 1.00
21 2,6-(C(CH ) ) ,4-C H
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
2
5
22 2,6-(C(CH
)
)
)
)
)
)
)
)
,4-CHO
23
24
25
2,6-(C(CH
2,6-(C(CH
2,6-(C(CH
In the present study, we examine the activation of
caspases and subsequent apoptosis of a library of
phenols against a mouse leukemia cell line. These
results are then compared with their corresponding
cytotoxicities in the same cell line to determine if
apoptosis plays a major role in the overall cytotoxicity
of monophenolic compounds as delineated by the form
of the corresponding QSAR and the type and magnitude
of the important descriptors.
,4-CH
,4-OH
2
OCH
3
3 3 3
26 2,4,6-(C(CH ) )
27
28
29
2,6-(C(CH
2,6-(C(CH
2,6-(C(CH
3
3
3
)
)
)
3
3
3
)
)
)
2
2
2
,4-CN
,4-CH
2
OH
3
,4-COCH
30 2,6-(C(CH ) ) ,4-Br
31 2,6-(C(CH
3
3
3
3
3
3
3
3
2
2
2
2
)
)
)
)
)
)
,4-CH
,4-NO
,4-OCH
3
(BHT)
3
3
3
3
2
3
4
5
2,6-(C(CH
2,6-(C(CH
2,6-(OCH
2,4,6-(OCH )
3 3
2
3
3
)
2
2.70 2.68 3.07 -0.02 1.00
2.20 2.68 3.07 -0.04 1.00
2.60 2.72 3.07 -0.25 1.00
3.10 2.90 3.07 -1.25 1.00
2.80 2.78 3.07 -0.57 1.00
2.50 2.85 3.07 -0.99 1.00
Results
36 2,6-(OCH ) ,4-CHdCHCHO
3
3
3
2
2
2
37
38
2,6-(OCH
2,6-(OCH
)
)
,4-NH
,4-COCH
2
The heterogeneous set of phenols in Table 1 includes
mono-, di-, and trisubstituted phenols with a large
number of sterically hindered derivatives. The range in
apoptotic activity is around 3 log units, and the highly
sterically hindered phenols generally have the highest
activities. With the data in Table 1, the following QSAR
equation was formulated, with one compound, 2-
iodophenol, omitted from the analysis.
3
39 2,6-(OCH ) ,4-NHCOCH
40 2,6-(OCH
3
3
3
2
2
2
3
)
)
,4-CH
,4-CHO
3
2.40 2.57 3.07
0.54 1.00
41
42
43
2,6-(OCH
2-NH
2-C(CH
2.90 2.80 3.07 -0.67 1.00
2.29 1.73 1.97 -1.23 1.00
2
3
)
3
2.42 2.42 3.17
2.38 2.50 3.17
1.87 1.48 2.04
2.38 1.49 2.15
2.11 2.59 3.17
2.16 2.83 3.49
1.98 1.00
1.53 1.00
0.56 1.00
1.12 1.00
1.02 1.00
1.55 1.00
3 2
44 2-CH(CH )
4
4
4
4
5
6
7
8
2-CH
3
b
2-I
2-C
2-C
2
H
H
5
7
3
Induction of caspase-mediated apoptosis
in L1210 cells by X-phenols:
49 2-NH
2
,4-NO
,4-CH
2
2.46 1.78 1.97 -1.51 1.00
2.09 1.63 1.97 -0.67 1.00
2.41 2.68 3.07 -0.02 1.00
5
5
0
1
2-NH
2-OCH
2
3
3
1
52 2-C(CH
)
, 4-OCH
(BHA)
2.71 2.42 3.17
1.96 1.00
3
3
3
log / ) (1.06 ( 0.12)B5 + (0.33 ( 0.20)B5 -
I50
2
3
a
Calculated using QSAR 4 b Not included in the derivation of
(
0.18 ( 0.09)π2,4 - (0.92 ( 0.46) (4)
QSAR 4
2
2
n ) 51, r ) 0.886, s ) 0.349, q ) 0.866,
the steric descriptor B52 while the hydrophobicity of the
ortho and para substituents as well as B53 accounts for
F
) 121.72
,47
3
5
% and 2%, respectively, of the variance in the data.
In eq 4, I50 represents the concentration of X-phenol
that induces caspase-mediated apoptosis by 50%. B52
is Verloop’s sterimol descriptor and is a measure of the
width of the larger substituent in the ortho position,
while B53 represents the width of the larger substituent
in the meta position. In most cases, ambiguity in
position assignment is minimized because of the sym-
metrical nature of the substitution patterns. However,
in the case of ortho monosubstituted analogues, e.g.,
This strong dependence on mostly steric terms suggests
that the phenols may be interacting with critical apo-
ptogenic proteins.
Equation 4 accounts for 89% of the variance in the
2
2
data. The cross-validated r (q ), which is high (0.87),
was obtained by using the leave-one-out (LOO) proce-
dure.14 Further validation of eq 4 was also carried out
by using Y-randomization. The biological data were
randomly shuffled five times and the following statisti-
15
2
2
-methoxyphenol, the methoxy substituent is placed in
cal values were obtained for the five runs: r ) 0.518,
2
2
2
2
2
the 2-position. The hydrophobic parameter π2,4 repre-
sents the sum of the hydrophobicity of substituents in
the para position and the bulkier ortho position. In this
model, 81% of the variance in the data is explained by
q ) 0.409; r ) 0.181, q ) 0.071; r ) 0.138, q )
2
2
2
2
-0.003; r ) 0.553, q ) 0.485; r ) 0.382, q ) 0.296.
2
2
These poor r and q values in the randomization test
ensures the robustness of eq 4. Analysis of the data in

7
236 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Selassie et al.
The monophenols that are most adept at inducing
apoptosis are clearly the multisubstituted ones with
bulky groups in the ortho positions. Thus, the most
potent phenolics are the 2,6-di-tert-butyl, 4-methoxy-
methyl and 2,6-dimethoxy, and 4-amino analogues
whose I50 values fall within the 1 mM range. At 0.1 mM,
however, all of these monophenolic compounds exhibit
minimal apoptosis (<15%).
To examine the relationship between caspase-medi-
ated apoptosis and cytotoxicity, we extended our exami-
nation of cell growth inhibition by substituted monophe-
nolics in L1210 cells. We expanded our study to include
six new 4-X-2,6-dimethoxy phenols. Once again, a
bifurcation in mechanism was observed depending on
the electron density in the aromatic ring. This phenom-
enon is illustrated in QSAR eqs 5 and 6, pertaining to
electron-releasing phenols (Table 2) and electron-at-
tracting phenols (Table 3), respectively. The outliers (∆
Figure 1. Induction of apotosis by 2-, 3-, and 4-methoxyphe-
nols.
Table 1 reveals that multisubstituted phenols are more
adept at inducing apoptosis than the monosubstituted
phenols. In the latter case, the following trend is
observed: ortho phenols . meta phenols . para phe-
nols. This disparity in apoptosis induction by positional
isomers is well illustrated in Figure 1.
The presence of steric terms (B52, B53) and a hydro-
phobic term (π2,4) suggests that large substituents, e.g.,
tert-butyl groups help to anchor the phenol on the
surface of a hydrophilic receptor. The small and negative
coefficient with the hydrophobic descriptor speaks to
this type of interaction. Despite the heterogeneity of the
substitutents’ electronic attributes, a bifurcation in
mechanism was not observed in the ability to induce
apoptosis. Thus electron-attracting and electron-releas-
ing phenols behave in a similar manner.
Cytotoxicity of electron-releasing phenols
in L1210 cells:
1
+
log /
) (-1.39 ( 0.19)σ - (0.28 ( 0.05)B5
+
2,6
ID50
(
0.16 ( 0.05)log P - (0.58 ( 0.24)I -
2
(1.04 ( 0.25)I
+ (3.90 ( 0.19) (5)
1
2
2
n ) 65, r ) 0.840, s ) 0.271, q ) 0.808,
F
) 62.07
5,59
Table 2. Cytotoxicity of Electron-Releasing Phenols in L1210 Cells
log 1/ID50
log 1/ID50
no.
X
obsd preda
σ+
B52,6 log P b
I
1
I
2
no.
X
obsd preda
σ+
B52,6 log P b
I
1
I
2
c
1
2
3
4
5
6
7
8
9
H
3.27 3.56
0.00 2.0
1.48
1.57
2.10
2.63
3.16
4.22
3.57
1.97
2.50
3.03
3.56
4.09
5.15
5.68
6.21
3.30
1.91
0.25
0.81
0.49
3.30
1.97
1.64
2.50
0.25
1.92
1.71
1.32
2.45
0.88
0.62
3.20
2.03
2.70
3.38
0.39
2.98
0
0
38 2-OC
2
H
5
3.25 4.08 -0.81 4.36
3.50 3.45 -0.60 4.61
5.03 4.99 -1.23 2.93
3.02 2.93 -0.62 4.08
3.86 3.91 -1.56 6.14
3.20 3.44 -0.93 4.08
3.85 3.60 -0.52 6.34
(BHT) 4.04 4.11 -0.83 6.34
3.26 3.47 -0.60 6.34
3.25 3.50 -0.56 6.34
3.90 4.25 -0.78 6.34
3.73 2.74 -0.57 5.21
1.85
0.18
1.38
2.37
1.11
2.87
4.93
5.43
3.43
3.93
6.75
3.65
5.96
6.09
5.03
3.70
2.42
2.36
3.08
2.02
4.23
1.07
1.01
1.00
1.01
1.61
0
0
4-OCH
3
4.48 4.08 -0.78 2.0
4.64 4.79 -0.81 2.0
4.85 4.90 -0.83 2.0
5.20 4.95 -0.81 2.0
5.50 5.12 -0.81 2.0
4.97 4.59 -0.50 2.0
3.85 4.07 -0.31 2.0
3.86 4.14 -0.30 2.0
4.04 4.21 -0.29 2.0
4.33 4.29 -0.29 2.0
4.47 4.38 -0.29 2.0
4.49 4.54 -0.29 2.0
4.62 4.63 -0.29 2.0
4.75 4.71 -0.29 2.0
4.09 4.21 -0.26 2.0
3.83 3.73 -0.07 2.0
5.09 5.17 -1.30 2.0
4.59 4.73 -0.92 2.0
3.73 4.24 -0.60 2.0
3.88 3.99 -0.10 2.0
3.54 3.74 -0.07 2.0
4.11 3.81 -0.16 2.0
3.71 3.82 -0.07 2.0
4.11 3.59 -0.16 2.0
3.52 2.73 -0.31 3.04
3.20 3.60 -0.07 2.35
3.78 3.45 -0.78 4.07
3.75 3.52 -0.30 4.17
4.92 4.48 -0.92 2.93
5.16 4.96 -1.30 2.97
4.00 3.58 -0.26 4.17
3.70 3.84 -0.60 4.26
3.50 3.53 -0.28 4.17
3.90 3.30 -0.30 5.45
2.70 2.97 -0.04 3.70
3.49 3.50 -0.29 4.49
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
39 2-NHCONH
2
0
0
1
0
1
0
0
0
0
0
1
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
1
1
1
1
1
1
0
0
0
0
1
0
0
4-OC
4-OC
4-OC
4-OC
4-OC
2
3
4
6
6
H
H
H
H
H
5
40 2-OH,4-CH
41 2,6-(CH
3 2
42 2,6-(OCH )
3
7
3 2
)
9
13
5
43 2,4,6-(CH
44 2,6-(C(CH
45 2,6-(C(CH
3 3
)
)
)
3
3
)
)
2
2
4-CH
3
3
3
-4-CH
3
4-C
4-C
4-C
4-C
4-C
4-C
4-C
2
3
4
5
7
8
9
H
H
H
H
H
H
H
5
2 5 2
46 2,6-(C H )
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
7
47 2,6-(CH(CH
3
)
2
)
)
2
9
48 2,4,6-(C(CH )
3
3
3
c
11
15
17
19
3 3 3
49 2-C(CH ) -6-CH
50 2,6-(C(CH
51 2,6-(C(CH
52 2,4-(C(CH
3
3
3
)
)
)
3
3
3
)
)
)
2
2
2
-4-C
2
H
5
3.91 4.18 -0.82 6.34
4.11 3.57 -0.37 6.34
4.24 4.23 -0.52 4.17
3.80 4.09 -0.57 4.17
3.04 3.23 -0.62 3.04
3.09 2.89 -0.38 3.04
3.46 2.70 -0.16 3.04
3.39 3.24 -1.09 3.04
3.80 4.16 -0.56 4.17
5.05 4.99 -2.34 6.14
3.10 3.25 -1.09 6.14
3.06 3.24 -1.09 6.14
-4-Br
4-C(CH
4-F
4-NH2
4-OH
4-NHCOCH
3-C(CH
3-CH
3-N(CH
3-C
3-NH
2-CH
2-F
2-OCH
2-C
2-OH
2-NH
2-C(CH
2-SCH
2-CH(CH
3
)
3
53 2-C(CH
3
)
3
-4-CH
3
(BMP)
54 2,4-(CH )
3 2
55 2-CH
56 2-CH
57 2-CH
3
3
3
-4-F
-4-Br
c
3
-4-OCH
3
3
)
3
58 2-C(CH
3
)
3
-4-C
3 3
-4-CHO
2
H
5
3
59 2,4,6-(OCH )
3
)
2
60 2,6-(OCH
61 2,6-(OCH
62 2,6-(OCH
63 2,6-(OCH
64 2,6-(OCH
65 2,6-(OCH
3
3
3
3
3
3
)
)
)
)
)
)
2
2
2
2
2
2
2
H
5
-4-COCH
3
2
-4-CHCHCHO 4.25 4.12 -1.72 6.14
c
c
3
-4-CH
-4-NH
-4-NHCOCH
(BHA)
,4-OCOCH
,4-CH
cOCH
,4-OH
,4-CH
3
2.89 4.42 -1.87 6.14
5.52 5.53 -2.86 6.14 -0.09
2
c
3
3
3.99 4.60 -2.16 6.14
3.82 4.10 -1.04 4.17
4.35 3.79 -0.71 6.34
3.93 3.64 -0.57 6.34
4.16 4.77 -1.44 6.34
3.88 3.49 -0.56 6.34
3.71 4.11 -1.30 6.34
4.06 4.25 -0.99 6.34
3.94 4.12 -0.90 6.34
0.15
3.30
4.46
4.72
4.26
3.89
5.03
4.89
4.91
2
H
5
66 2-C(CH
3
)
3
,4-OCH
3
67 2,6-(C(CH
68 2,6-(C(CH
69 2,6-(C(CH
70 2,6-(C(CH
71 2,6-(C(CH
72 2,6-(C(CH
73 2,6-(C(CH
2
2
2
2
2
2
2
)
)
)
)
)
)
)
3
3
3
3
3
3
3
)
)
)
)
)
)
)
2
2
2
2
2
2
2
3
2
2
3
3 3
)
3
2
OH
3
)
2
,4-OCH
,4-CHO
3
d
c
2-CH
2-CH
2-C
2
3 2
CH (CH )
OH
d
2
,4-COCH
3
3
H
7
a
b
c
d
+
P
r
e
d
i
c
t
e
d
u
s
i
n
g
e
q
5
.
C
a
l
c
u
l
a
t
e
d
u
s
i
n
g
C
l
o
g
P
v
e
r
s
i
o
n
4
.
2
3
.
N
o
t
i
n
c
l
u
d
e
d
i
n
d
e
r
i
v
a
t
i
o
n
o
f
Q
S
A
R
5
.
A
p
p
r
o
x
i
m
a
t
e
l
y
v
a
l
u
e
s
o
f
σ
calculated according to ref 34.

QSAR Study of Phenolic Compounds
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7237
Table 3. Cytotoxicity of Electron-Attracting Phenols versus
L1210 Cells
Table 4. Cytotoxicity of Phenolic Compounds versus HL-60
and MCF-7 Cells
1
log /ID50
HL-60
MCF-7
1
1
log /ID
log /ID
preda
log P b
50
50
no.
X
obsd
B52
no.
X
obsd preda obsd predb log P
σ+
1
2
3
4
5
6
7
8
9
H
3.27
2.48
3.45
3.86
2.50
3.08
4.29
4.20
3.44
3.48
2.65
3.87
3.71
3.82
3.11
3.46
3.46
3.22
3.30
3.34
3.44
3.95
3.22
4.90
3.49
3.14
4.68
3.32
2.67
3.53
4.12
2.49
3.30
3.89
3.97
3.39
3.53
2.77
3.89
3.37
3.97
3.39
3.57
2.94
3.46
3.21
3.10
3.53
3.58
3.58
4.83
3.47
3.25
4.69
1.48
0.33
1.85
2.89
0.01
1.44
2.48
2.63
1.60
1.85
0.49
2.48
1.57
2.63
1.60
1.91
0.81
2.15
1.60
1.85
2.35
2.54
2.80
5.31
2.30
1.90
5.05
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.80
1.60
2.44
1.95
2.15
2.61
3.17
2.04
2.04
3.17
4-CONH2
4-NO2
4-I
1
2
3
4
5
6
7
8
9
H
3.05 3.06 2.56 2.78 1.48
0.00
1.97 -0.31
4-CH3
4-OCH3
3.35 3.34
3.60 3.53 3.52 3.55 1.57 -0.78
4-SO2NH2
4-CHO
4-(CH2)3CH3
4-OC3H7
4-NH2
diethylstilbesterol 4.00 3.89
equilenin
â-estradiol
3.52 3.67
3.70 3.78
3.56 -0.29
2.63 -0.83
c
4-Cl
4.50 3.54 4.10 4.06 0.25 -1.30
4-Br
4.96 -0.16
3.27 -0.42
3.78 -0.35
-0.30
4-CN
3.90 3.68
3.60 3.75
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
3-NO2
3-NHCOCH3
3-Cl
10 4-C2H5
11 4-OC2H5
12 4-OC6H5
2.91 3.07
3.70 3.58
3.99 3.27
3.11 3.03
-0.81
d
3-OCH3
3-Br
-0.50
13 4-C(CH3)3
-0.26
3-CN
a
Calculated using eq 7. Calculated using eq 8. c Not included
in the derivation of eq 7. Not included in the derivation of eq 8.
b
3-F
d
3-OH
2-Cl
The negative coefficients of the I1 (-1.04) and I2
(-0.58) variables are particularly interesting. They
highlight the diminished cytotoxicity of ortho methyl
and methoxy substituted phenols. In both cases, their
diminished cytotoxicities could be attributed to their
unbridled reactivities and subsequent indiscriminate
reaction with the surrounding aqueous medium. Their
enhanced reactivities are a function of their high σ
values and increased intermolecular hydrogen bonding
with water.
2-CN
2-NO2
2-Br
2-I
2-CF3
2,6-(C(CH3)3)2-4-NO2
2-CH3-4-NO2
2-CH3-4-COCH3
2,6-((CH3)3)2, 4-CN
+
a
Calculated using eq 6. b Calculated using ClogP, version 4.23.
Eight of the phenols (nos. 26, 35, 38, 49, 56, 63, 65,
and 69) in Table 2 were deemed to be outliers on the
basis of their deviations (2 SD). Ortho substitution
appears to be a common factor in this subset, particu-
larly when bulky ortho substituents flank the phenolic
group as is seen in four of the cases. The general
>
2 SD) are 2-CH3; 2-CH2CH(CH3)2; 2-OC2H5; 2-C(CH3)3,
6
4
-CH3; 2-CH3, 4-Br; 2,6-(OCH3)2, 4-CH3; 2,6-(OCH3)3,
-NHCOCH3; and 2,6-(C(CH3)3)2, 4-OH.
B52,6 represents the sum of the width of the substit-
uents in the ortho position, the two substituents that
flank the labile hydroxy group. Cytotoxicity decreases
as the width of these substituents increase. The negative
+
approach utilizing the additivity of σ may not provide
an accurate representation of electron density in these
instances. Reasons for the anomalous behavior of 2-eth-
oxyphenol were not apparent. Three phenols with
methyl groups in the ortho positions (nos. 26, 49, 56)
but not the para position (63) were much more active
than predicted.
With the data in Table 3, the following QSAR equa-
tion was generated for the inhibition of growth of L1210
cells by electron-attracting phenols.
+
coefficient with σ (-1.39) implies that highly electron-
releasing substituents (e.g., NH2, OCH3) enhance sta-
bilization of the phenoxy radical and increase subse-
quent cytotoxicity. I1 and I2 are indicator variables,
which pinpoint the unusual activities of the ortho
methyl and methoxy substituents, respectively. They
both have a deleterious effect on cytotoxicity as evi-
denced by their negative coefficients (-1.04 and -0.58).
+
The various descriptors σ , B52,6, I1, log P, and I2 make
the following sequential contributions to the variance
in the data: σ (17%), B52,6 (24%), I1 (19%), log P (18%),
and I2 (6%). The coefficient with the hydrophobic
descriptor is small and suggests a surface interaction
with a minimally hydrophobic receptor. The most cyto-
toxic phenols tend to be the unhindered ones with high
electron-releasing capability such as 2-NH2 (5.16); 4-NH2
Cytotoxicity of electron-attracting phenols
in L1210 cells:
+
1
log /
) (0.56 ( 0.11)log P - (0.30 ( 0.18)B5 +
2
ID50
(2.79 ( 0.22) (6)
2
2
n ) 27, r ) 0.848, s ) 0.233, q ) 0.812,
(5.09); 4-OC6H13 (5.50); and 2,6-(OCH3)2, 4-NH2, (5.52)
F
) 66.91
,24
2
and thus favored with enhanced radical-stabilizing
ability. In the hexyloxy case, the hydrophobicity of the
long chain also enhances cytotoxicity. BHT and 2,6-di-
tert-butyl-4-methylphenol (BMP) are moderately toxic
with IC50 values of 91.2 and 158.5 µM, respectively.
BHA is similar to BMP in terms of its cytotoxic
The log P parameter is of critical importance in
describing the cytotoxicity in this cell line. It alone
accounts for 84% of the variance in the data. Note that
the coefficient with the log P term is much greater than
that seen in eq 5. The negative coefficient with Verloop’s
potential, with an IC50 value of 151 µM. As in QSAR eq
B5 parameter suggests that steric hindrance plays a
2
+
1
, the strong dependence on σ is consistent with the
role, albeit a minor one in facilitating membrane
suggestion that radical stabilization effects are of
significance in the case of electron-releasing substitu-
ents.
penetration.
Table 4 includes a set of miscellaneous phenols whose
cytotoxicities were assessed versus HL-60 (human
1
6

7
238 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Selassie et al.
Table 5. Cytotoxicity of Phenolic Compounds versus CCRF
tal human acute lymphoblastic cells) sensitive and
resistant (CEM/VLB) to vinblastine. The phenols were
sequestered according to their electron densities, and
eqs 9-12 were formulated for their activities.
(
Sensitive) and CEM/VLB (Resistant) Cells
CCRF-CEM CEM/VLB
1
1
log /ID
log /ID
50
50
no.
X
obsd pred obsd pred log P
σ+
4.29 4.00 3.86 4.00b 2.89
a
Inhibition of growth of CCRF-CEM cells by
electron-attracting phenols:
1
2
3
4
5
6
7
8
9
4-I
a
b
4-NO
2
3.28 3.33 3.36 3.34b 1.85
a
4-SO
2
NH
2
2.31 2.14 2.11 2.16 0.01
a
b
1
4-CONH
2
2.23 2.35 2.38 2.36 0.33
a b
log /
) (0.64 ( 0.32)log P + (2.14 ( 0.62) (9)
ID50
4-Cl
3.45 3.74 3.89 3.74 2.48
c d
e
4-NH
2
H
4.61 4.74 1.94 5.25 0.25 -1.30
c d
2
2
4-OC
4-CH
6
13
5.34 5.10 5.37 4.33 4.22 -0.81
c d
n ) 5, r ) 0.929, s ) 0.265, q ) 0.768
3
3.82 3.71 4.05 3.61 1.97 -0.31
c d
4-C(CH
3
)
3
3.86 4.01 3.79 3.88 3.30 -0.26
c d
Inhibition of growth of CEM/VLB cells by
electron-attracting phenols:
1
1
1
1
1
1
0
1
2
3
4
5
4-H
3.27 3.10 2.65 2.80 1.48
0.00
c
d
4-OCH
3
4.41 4.32 4.36 4.51 1.57 -0.78
c
d
4-C
3
H
7
3.72 3.98 3.73 3.87 3.03 -0.29
c d
1
4-C(CH
3
)
2
C
6
H
4
-4′-OH 4.18 4.15 3.96 4.05 3.67 -0.29
log /
) (0.64 ( 0.15)log P + (2.15 ( 0.29) (10)
c
d
ID50
4-C
4-C
2
H
H
5
3.79 3.84 3.84 3.74 2.50 -0.30
c d
8
17
4.66 4.71 4.67 4.63 5.68 -0.29
2
2
n ) 5, r ) 0.983, s ) 0.124, q ) 0.949
a
b
c
Calculated using eq 9. Calculated using eq 10. Calculated
using eq 11. d Calculated using eq 12. Not included in the
e
Inhibition of growth of CCRF-CEM cells by
electron-releasing phenols:
derivation of eq 12.
promylocytic leukemia cell line) and MCF-7 (human
breast cancer cell line). Equations 7 and 8 were devel-
oped for the inhibition of growth of HL-60 and MCF-7,
respectively, by these phenols.
1
+
log /
) (0.28 ( 0.10)log P - (1.52 ( 0.39)σ +
ID50
(2.69 ( 0.40) (11)
2
2
n ) 10, r ) 0.933, s ) 0.176, q ) 0.807
Inhibition of growth of HL-60 by phenolic
compounds:
Inhibition of growth of CEM/VLB cells by
electron-releasing phenols:
1
+
log /
) (0.21 ( 0.13)log P - (0.57 ( 0.53)σ +
ID50
1
+
log /
) (0.29 ( 0.14)log P - (2.16 ( 0.74)σ +
ID50
(
2.75 ( 0.48) (7)
(2.37 ( 0.51) (12)
2
2
n ) 8, r ) 0.816, s ) 0.152, q ) 0.626,
outlier ) 4-NH2
2
2
n ) 9, r ) 0.935, s ) 0.219, q ) 0.859,
outlier ) 4-NH2
Inhibition of growth of MCF-7 by phenolic
compounds:
Once again, a similar dependence of phenolic cyto-
toxicity on only hydrophobicity is seen in both sensitive
and resistant cells in the case of electron-attracting
phenols, which suggests that nonspecific toxicity at the
membrane level is operating in both cell lines. A greater
difference in response is seen in the case of electron-
releasing phenols. Cytotoxicity is dependent on the
lipophilic and electronic character of the phenols; the
electron-releasing substituents play a greater role in
stabilizing the reactive intermediate in resistant cells.
CEM/VLB is a Pgp overexpressing resistant subline,17
and the sensitivity of this line to electron-releasing
phenols is enhanced by the electron-rich character of
1
+
log /
) (-1.16 ( 0.31)σ + (2.65 ( 0.23) (8)
ID50
2
2
n ) 6, r ) 0.964, s ) 0.120, q ) 0.916,
outlier ) 4-OC H
6
5
In this size-limited set, growth inhibition is also
minimally dependent on hydrophobicity and mostly
affected by electron density as seen in eqs 1 and 5. The
coefficient with the hydrophobic term is similar, but the
coefficient with σ+ is lower. However, the 95% confi-
dence interval is also much larger. This is in keeping
with the smaller data set and the subsequent limited
+
the substituents. The coefficient with the σ term
+
range in σ values. It also appears that HL-60 is less
(-2.16) is a measure of the susceptibility of cytotoxicity
to electronic effects. It has been determined that CEM/
VLB has a more active mitochondrial electron transport
chain than its parental CCRF-CEM cell line.18 This
could account for the close interaction between the easily
available radical intermediate and the electron trans-
port chain.
susceptible to electronic perturbations than the L1210
cell line. This data set does include three estrogenic
phenols.
The small number of data points precluded the
examination and inclusion of other descriptors. A strong
dependence of growth inhibition on electron-releasing
capabilities of the phenols is observed; a large negative
coefficient with σ+ is seen. Equations 1, 5, 7, and 8
reveal a recurring theme in these cytotoxicity studies,
Discussion
The QSAR for apoptosis contrasts sharply to the
QSAR for cytotoxicity (Table 6). A significant difference
+
a strong dependence on σ and hence radicalization of
+
the phenol to its ensuing phenoxy radical. The use of
BDE and their excellent correlations with inhibitory
potencies (eq 2) validates the importance of hydrogen
abstraction in this critical ligand-receptor interaction.
Table 5 includes a limited set of X-phenols whose
cytotoxicities were assessed versus CCRF-CEM (paren-
involves the lack of the electronic parameter σ . Thus,
apoptosis must include a distinct step that focuses on
the interactions of the phenolic moiety with a critical
receptor such as a mitochondrial protein, procaspase,
or cytochrome c. Mitochondrial membrane polarization
(MMP) is a key step in the intrinsic pathway that

QSAR Study of Phenolic Compounds
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7239
Table 6. Summary of Critical Variables in Eqs 4-12
descriptor coefficients
QSAR
equation no.
n
cell line
L1210
compd
all phenols
steric
hydrophobic
electronic
other
intercept
Apoptosis
0.33B52
.06B52
4
51
-0.18π2.4
-0.92
1
Cytotoxicity
5
63
L1210
electron-releasing phenols
-0.28B52,6
0.16 log P
-1.40σ+
-0.61I2
3.88
-
1.05I1
6
7
8
9
0
1
2
27
8
L1210
electron-attracting phenols
X-phenols
-0.30B52
0.56 log P
0.21 log P
2.79
2.75
2.65
2.14
2.15
2.69
2.37
+
HL-60
-0.57σ
+
6
MCF-7
X-phenols
-1.16σ
5
CCRF-CEM
CEM/VLB
CCRF-CEM
CEM/VLB
electron-attracting phenols
electron-attracting phenols
electron-releasing phenols
electron-releasing phenols
0.64 log P
0.64 log P
0.28 log P
0.29 log P
1
1
1
5
+
10
9
-1.52σ
+
-2.16σ
characterizes apoptosis.19 Induction of MMP is sufficient
to bring on apoptosis or necrosis. Reactive oxygen
species can trigger MMP, which involves drastic alter-
ations in cell metabolism and the release of apoptogenic
proteins that initiate caspase-dependent as well as
2
0,21
caspase-independent destructive pathways.
Yu et al. in conclusive studies have shown that the
cytotoxicity of BHA is due to the induction of apoptosis
as a direct consequence of release of cytochrome c and
22
subsequent activation of caspases. Pretreatment of rat
hepatocytes with antioxidants such as N-acetylcysteine
and ascorbic acid had little effect on apoptosis, suggest-
ing that induction of apoptosis by BHA appears to be
independent of formation of reactive metabolites. Pre-
liminary results in my laboratory with N-acetylcysteine
and ascorbic acid support these findings in L1210 cells.23
QSAR eq 4, pertaining to caspase-mediated apoptosis,
is also supportive of this view because it lacks the
+
ubiquitous σ term that predominates in cytotoxicity
QSAR. The mostly steric terms suggest that the phenols
may be interacting with apoptotic proteins released by
the direct action of phenols on mitochondria. The
X-phenols in this study are moderately hydrophobic and
will have the ability to partition into the hydrophobic
phase of the membrane lipids, which may interfere with
membrane integrity particularly in the case of the bulky
analogues. Bulky substituents in the ortho and meta
positions of relatively small hydrophilic phenols may
enhance apoptosis by binding to a hydrophilic receptor
like cytochrome c or maybe even the procaspases
directly, resulting in caspase induction or activation.
Figure 2. Contrasting behavior of BHA and BHT.
1
An analysis of the difference (∆ log /C) between the
cytotoxicity and apoptosis values for the phenols in this
study revealed a number of compounds with values less
than 1.70. They included phenols 11, 21, 30, 34, 36, 38,
39, 41, 43-48, 51, and 52. Phenols 43-48 and 51 are
ortho mono-substituted phenols, which indicates that
at cytotoxic concentrations, some apoptosis may ensue
(∼10%). Phenols that could induce considerable apo-
ptosis at cytotoxic concentrations include 2,6-dimeth-
oxyphenol (1.16), 2,6-dimethoxy-4-acetylphenol (0.26),
2,6-dimethoxy-4-formylphenol (0.20), 2-isopropylphenol
(1.12), and BHA (1.11). However at its IC50, the level of
apoptosis induced by BHA is observed to be minimal,
around 15-20%. BHT does not induce apoptosis at
2
4
Saito et al. have shown that caspases 3, 8, and 9
can be activated to varying degrees by BHT, BHA, and
BMP in HL-60 cells. The concentrations that induced
DNA fragmentation were between 0.1 and 0.2 mM for
both BHT and BHA. Exact I50 values were not available.
In our assays, using HL-60 cells, the I50 values of BHT
and BHA were assessed after 16 h and determined to
be 3.18 and 1.38 mM, respectively. Thus, in both cell
lines, L1210 and HL-60, BHA was found to be more
apoptogenic than BHT. Recently, Okubo et al. assessed
the cytotoxicity of BHA in human monocytic leukemia
U937 cells. It caused nuclear condensation and frag-
mentation, structural damage in mitochondria, reduc-
tion in mitochondrial transmembrane potential, inter-
nucleosomal DNA cleavage, and induction of caspase
1
cytotoxic doses (as seen by its ∆ log /C ) 1.92), which
translates into a concentration differential of 80. Figure
2 illustrates the stark differences in behavior of BHA
and BHT at a range of concentrations.
+
The significant presence of σ in eqs 5, 7, and 8
underscores the importance of phenoxy radicals in
mediating the ensuing inhibition of cell growth. It is well
established that electron-donor groups enhance radical
stabilization via an increased through-resonance ef-
2
5
activity that are hallmarks of apoptosis. From their
results, it was established that cytotoxicity induced by
BHA was mostly attributable to apoptosis phenomena.

7
240 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Selassie et al.
fect.26 Excellent correlations between BDE (O-H) and
reactive species react rapidly via nonenzymatic Michael
additions to critical nucleophiles, to induce toxicity.
+
Brown’s σ have been obtained for para-substituted
1
6,27,28
phenols.
Bordwell et al. obtained a linear plot of
BDEs of the O-H bonds for 14 para-substituted phenols
Conclusion
+
+
16
versus σ with a F value of 7.14. Our similar analysis
In a comparison of the concentrations of phenols
utilized to induce apoptosis after 12 h and cytotoxicity
after 48 h, we see marked differences. Apart from the
simple monosubstituted phenols, most of the multisub-
stituted phenols mediate apoptosis at the millimolar
+
of a set of 28 phenols yielded a F of 6.12 ( 0.47.
The negative sign and magnitude of the steric term
B52,6) indicate that extensive crowding of the O-H
(
moiety by bulky ortho substituents leads to a small
decrease in cytotoxicity. Perhaps the steric repulsion
between the phenolic OH and the adjacent substituents
with significant widths destabilizes the parent phenol
-3
level (10 mol). This contrasts sharply with the cyto-
toxic concentrations that hover around the 0.1 milli-
-4
molar range (10 mol). This difference indicates that
2
9
as well as the ensuing radical.
at critical cytotoxic concentrations, apoptosis is mostly
nonexistent or minimal at best (<5%). There are four
phenols that do not fall under this umbrella, and they
include 2,6-dimethoxy-4-acetylphenol and 2,6-dimethoxy-
The indicator variable I1 eludes to the observed
diminished cytotoxicity of methyl-substituted phenols.
This unusual behavior (considering its electron-donating
capability) raises a question concerning the identity of
the reactive site and the substituent. Should such
compounds be considered as substituted phenols or
substituted toluenes, both of which are subject to
1
4-formylphenol, whose log / values are observed to be
C
similar in the cytotoxic and apoptosis assays. This
behavior suggests that their cytotoxicity may be at-
tributed to their apoptosis-inducing ability. The other
two phenols, 2,6-dimethoxyphenol and 2-tert-butyl-4-
methoxyphenol (BHA), show minimal (<15%) apoptosis
at cytotoxic-inducing concentrations.
30
hydrogen abstraction? There exists many examples of
hydrogen abstraction from the benzylic carbon of tolu-
+
31
ene that are well correlated by σ . The various
+
+
coefficients with the σ term (the F values) range from
2.53 to -0.32, depending on the solvent, reagent, and
The distinct differences between the eqs 4 and 5
suggest that cytotoxicity may not be a direct result of
-
temperature. It is also well established that benzylic
carbon atoms are susceptible to oxidation, easily forming
the corresponding alcohol metabolite.32 These observa-
tions suggest that benzylic hydroxylation could well be
occurring in the L1210 cell line, thus providing another
avenue of metabolism for X-alkyl phenols.
+
apoptosis. The lack of a σ term in the apoptosis QSAR
downplays the importance of a reactive intermediate
such as a phenoxy radical in its occurrence. The pres-
ence of hydrophobic and steric terms in eq 4 suggests
that apoptosis is a receptor-mediated process with
mitochondrial proteins or caspases being the likely and
ultimate targets. Preliminary results with a few phenols
indicate that at apoptosis-inducing concentrations, BHT,
BHA, and 2-amino-p-cresol do disrupt mitochondrial
transmembrane potential in L1210 cells.37
In summary, data on an extensive series of substi-
tuted monophenols in different murine and human cells
indicate that apoptosis occurs at high concentrations in
contrast to cytotoxicity, which in most cases is not
mediated by apoptotic phenomena. Cytotoxicity is an
end result of complex interactions of electronic, solva-
tion, steric, and hydrophobic factors. The ability to form
a reactive phenoxy species underscores the importance
of cleavage of the O-H bond in their chemical reactivi-
ties and subsequent cytotoxic activities.
The indicator variable I2 pinpoints the slightly aber-
rant behavior of methoxy-substituted phenols, mostly
those of the 2,6 variety. Using BDE data, various groups
have stressed that hydrogen bonding is responsible for
the stabilizing behavior of methoxy (hydrogen bond
acceptors) phenols in what is deemed to be the “toward
conformer” (where the hydrogen bond points toward the
2
6,29,33
methoxy moiety).
Thus, intramolecular hydrogen
bonding could contribute to reduced delocalization of the
critical radical that mediates cytotoxicity. Hence, the
utility of more lipophilic analogues of 2,4,6-trimethoxy-
phenol and 4-amino-2,6-dimethoxyphenol as potential
effective antioxidants merits further attention. It must
be noted that although quantum mechanical and em-
pirical methods have been used extensively to assess
the kinetic and thermodynamic stabilities of the phe-
nolic bond, discrepancies in the published values for
phenol itself have led to a recent study by Bosque and
Sales, who advocated the use of molecular structural
descriptors to calculate more accurate bond dissociation
Experimental Procedures
Melting points were determined on an electrothermal melt-
ing point apparatus (MEL-TEMP II with digital thermometer).
1
13
H NMR and C NMR spectra were recorded on a Bruker DPX
400 MHz NMR spectrometer with TMS as the internal
standard; chemical shifts are given in δ (ppm). Thin-layer
chromatography was performed on silica gel plates (silica gel
IB-G Baker). Chemical elemental analysis was carried out by
Desert Analytics (Tucson, AZ).
3
4
energies of phenols.
Phenolic compounds with ortho or para alkyl substit-
uents may be easily oxidized via an initial phenoxy
radical to a more reactive quinone methide, which can
subsequently alkylate cellular proteins and/or DNA to
induce cytotoxicity.35 The data set in Table 2 contains
Chemicals. Most of the phenols and estrogenic phenols are
commercially available. The synthesis of some of the ortho-
12,38
substituted phenols have been previously reported.
2,6-
3
6
2
7 of this class of phenols. Bolton et al. have shown
Dimethoxyphenol, 2,6-dimethoxy-4-acetophenol, 2,6-dimethoxy-
-formylphenol, and 3,5-dimethoxy-4-hydroxycinnamaldehyde
4
that the presence of bulky alkyl groups in the ortho
positions provides little or no stabilizing influence on
the oxo group of the quinone methide and also dimin-
ishes its interactions with the aqueous medium, unlike
its charged resonance counterpart, in more accessible
quinone methides. Thus, in the neutral form these
were purchased from Aldrich Chemical Co. (Milwaukee, WI),
while 2,6-dimethoxy-4-methylphenol was purchased from Alfa
Aesar Chemical Co. (Ward Hill, MA). 2,4,6-Trimethoxyphenol
was obtained from Pfaltz & Bauer (Waterbury, CT). Two other
phenols, 4-amino-2,6-dimethoxyphenol and 4-acetamido-2,6-
dimethoxyphenol were synthesized as follows.

QSAR Study of Phenolic Compounds
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23 7241
Syntheses. 4-Amino-2,6-dimethoxyphenol Hydrochlo-
ride. 4-Hydroxy-3,5-dimethoxyazobenzene (1.3 g, 0.005 mol),
obtained from the reaction of 2,6-dimethoxyphenol with ben-
zenediazonium chloride, was dissolved in 25 mL of hot
ethanol, and a solution of sodium hydrosulfite (3.5 g in 18 mL
of water) was added slowly. The reaction mixture was heated
at reflux for 2 h and then immediately filtered. The filtrate
was concentrated to 10 mL and then cooled. The product so
obtained was filtered, washed with water, and dried. It gave
dilutions were made in unsupplemented Dulbecco’s medium.
The procedures utilized to determine the IC50 values for the
4
2
various phenols have been previously described.
3
9
Determination of Caspase Activity. A fluorimetric,
homogeneous caspase assay (Roche Molecular Biochemicals,
Mannheim, Germany) was utilized for the quantitative in vitro
determination of caspase activity in L1210 cells. Cells (4 ×
4
1
3
0 cells/mL) were seeded in 96-well plates and incubated at
7 °C for 12 h in the presence or absence of varying concentra-
8
00 mg of crude 4-amino-2,6-dimethoxyphenol, which was
tions of each substituted phenol. Then the fluorescent caspase
substrate DEVD-R110 (Asp-Glu-Val-Asp-rhodamine 110) pre-
diluted in incubation buffer was added and the microplates
were incubated at 37 °C for 2 h. The degree of fluorescence
was measured at 521 nm after excitation at 490 nm. The
concentration of X-phenol that induced caspase activity by 50%
dissolved in 100 mL of ethanol, and dry hydrogen chloride gas
was passed through it. The solvent was then evaporated, and
the residue was washed with ether and dried. It was crystal-
lized from ethanol to yield the hydrochloride of 4-amino-2,6-
1
dimethoxyphenol (740 mg, 71.49%), mp 214-215 °C. H NMR
δ
H
(400 MHz, DMSO-d
6
) 3.82 (6 H, s, 2 × OCH
3
), 6.62 (2 H, s,
). C NMR δ (400
), 101.0 (C3 and C5), 122.9
C1), 135.3 (C4), 148.6 (C2 and C6). Anal. Found (%): C, 46.51;
H, 5.99; N, 6.85; Cl, 16.97. Calcd for C Cl: C, 46.7; H,
.84; N, 6.81; Cl, 17.27.
-Acetamido-2,6-dimethoxyphenol. 4-Amino-2,6-dimeth-
(I
50) was calculated and used to derive the appropriate QSAR
equation.
+
13
Ph), 8.70 (1 H, s, OH), 9.96 (3 H, s, NH
MHz, DMSO-d
) 56.5 (2 × OCH
3
C
6
3
QSAR Analysis. The CQSAR suite of programs (BioByte,
(
4
3
Inc., Claremont, CA) was used to derive the various models.
H12NO
8 3
P represents the octanol-water partition coefficients of the
phenols. In this study, calculated log P (CLogP) values were
5
4
+
used to represent hydrophobicity. σ is Brown’s refinement of
oxyphenol (760 mg, 4.5 mmol), as prepared above, was dis-
solved in a mixture of glacial acetic acid (3 mL) and acetic
anhydride (3 mL), and the mixture was stirred overnight at
room temperature. The solvent was evaporated at reduced
pressure, and the residue was dissolved in ethyl acetate. The
ethyl acetate layer was washed with water, dried over mag-
nesium sulfate, and finally filtered. The filtrate was evaporated
to provide the crude product, which was crystallized from
ethanol to give 4-acetamido-2,6-dimethoxyphenol (660 mg,
44
the Hammett electronic constant σ. In all equations, n
represents the number of data points, r is the correlation
coefficient, s is the standard deviation of the regression
2
2
equation, and q represents the cross-validated r . The 95%
confidence intervals for the terms in the equations are listed
2
2
in parentheses. The cross-validated r (q ) is obtained by using
14
the leave-one-out (LOO) procedure of Cramer et al.
4
0 1
References
6
9.62%), mp 157-159 °C (lit. mp 141 °C). H NMR δ
) 2.01 (3 H, s, CH
), 3.72 (6 H, s, 2 × OCH
.92 (2 H, s, Ph), 8.08 (1 H, s, OH), 9.68 (1 H, s, NH). C NMR
(400 NHz, DMSO-d ) 24.3 (CH ), 97.8 (C3
), 56.2 (2 × OCH
C5), 131.3 (C1), 131.7 (C4), 148.0 (C2 & C6), 168.0 (CO).
H
(400
3
MHz, DMSO-d
6
3
),
(1) Aravind, L.; Dixit, V. M.; Koonin, E. V. The Domains of Death:
13
6
Evolution of the Apoptosis Machinery. Trends Biochem. Sci.
1
999, 24, 47-53.
δ
C
6
3
3
(
(
(
2) Denault, J.-B.; Salvesen, G. S. Caspases: Keys in the Ignition
&
of Cell Death. Chem. Rev. 2002, 102, 4489-4499.
3) Thornberry, N. A.; Lazebnik, Y. Caspases: Enemies Within.
Science (Washington, D.C.) 1998, 281, 1312-1316.
Anal. Found (%): C, 56.69; H, 6.30; N, 6.62. Calcd (%) for
: C, 56.87; H, 6.20; N, 6.63.
Evaluation of Cytotoxicity. L1210 (murine leukemia
cells), HL-60 (human promylelocytic leukemia cells), CCRF
drug sensitive parental human acute lymphoblastic cells), and
CEM-VLB (multidrug-resistant subline of CCRF-resistant to
10 4
C H13NO
4) Belmokhtar, C. A.; Hilion, J.; Dudognon, C.; Fiorentino, S.;
Flexor, M.; Lanotte, M.; Segal-Bendirdjian, E. Apoptosome-
Independent Pathway for Apoptosis. J. Biol. Chem. 2003, 278,
(
2
9571-29580.
(
(
(
(
5) Zou, H.; Li, Y.; Liu, X.; Wang, X. An APAF-1.Cytochrome c
Multimeric Complex Is a Functional Apoptosome That Activates
Procaspase-9. J. Biol. Chem. 1999, 274, 11549-11556.
6) Talapatra, S.; Thompson, C. B. Growth Factor Signaling in Cell
Survival: Implications for Cancer Treatment. J. Pharmacol.
Exp. Ther. 2001, 298, 873-878.
-
1
1
00 ng mL vinblastine) cells were maintained in asynchro-
nous logarithmic growth at 37 °C in RPMI medium with
L-glutamine supplemented with 10% (v/v) FBS. All stock
solutions and dilutions were made in unsupplemented RPMI
medium.
7) Johnstone, R. W.; Luefli, A. A.; Lowe, S. W. Apoptosis: A Link
Between Cancer Genetics and Chemotherapy. Cell 2002, 108,
4
Cell cultures were seeded at (2-5) × 10 cells/mL in
1
53-164.
duplicate for each inhibitor concentration in a 96-well micro-
titer plate (180 µL/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 eight concentrations. After 48 h of continuous
drug exposure, the cells were counted by using the CyQUANT
GR assay kit from Molecular Probes. For this purpose the
media was removed from the plates, which were then frozen
at -80 °C for a minimum of 1 h. The cells were thawed at 37
8) Muller, M.; Strand, S.; Hug, H.; Heinemann, E. M.; Walczak,
H.; Hofmann, W. J.; Stremmel, W.; Krammer, P. H.; Galle, P.
R. Drug-Induced Apoptosis in Hepatoma Cells Is Mediated by
the CD95 (APO-1/Fas) Receptor/Ligand System and Involves
Activation of Wild-Type p53. J. Clin. Invest. 1997, 99, 403-413.
(9) Muller, M.; Wilder, S.; Bannasch, D.; Israeli, D.; Lehlbach, K.;
Li-Weber, M.; Friedman, S. L.; Galle, P. R.; Stremmel, W.; Oren,
M.; Krammer, P. H. p53 activates the CD95 (APO-1/Fas) Gene
in Response to DNA Damage by Anticancer Drugs. J. Exp. Med.
1998, 188, 2033-2045.
°
C, and 200 µL of CyQUANT GR dye/cell lysis buffer was
added to each well. The plates were incubated for 5 min at 37
C, and their fluorescence was measured using a Cytofluor II
(10) Stich, H. F. The Beneficial and Hazardous Effects of Simple
Phenolic Compounds. Mutat. Res. 1991, 259, 307-324.
(
11) Garg, R.; Kurup, A.; Hansch, C. Comparative QSAR: On the
°
Toxicology of the Phenolic OH Moiety. Crit. Rev. Toxicol. 2001,
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 was
determined. The CyQUANT GR assay measures the ability of
CyQUANT GR dye to bind to cellular nucleic acids. Cytotox-
icity is expressed as the concentration of the phenol (IC50) that
causes a 50% reduction in fluorescence compared with the
3
1, 223-245.
(
12) Selassie, C. D.; Shusterman, A. J.; Kapur, S.; Verma, R. P.;
Zhang, L.; Hansch, C. On the Toxicity of Phenols to Fast
Growing Cells. A QSAR Model for a Radical-Based Toxicity. J.
Chem. Soc., Perkin Trans 2, 1999, 2729-2733.
(
13) Selassie, C. D.; DeSoyza, T. V.; Rosario, M.; Gao, H.; Hansch,
C. Phenol Toxicity in Leukemia Cells: A Radical Process? Chem.
Biol. Interact. 1998, 113, 175-190.
14) Cramer, R. D.; Bunce, J. D.; Patterson, D. E. Cross Validation,
Bootstrapping and Partial Least Squares Compared with Mul-
tiple Regression in Conventional QSAR Studies. Quant. Struct.-
Act. Relat. 1988, 7, 18-25.
15) Tropsha, A.; Grammatica, P.; Gombar, V. K. The Importance of
Being Earnest: Validation Is the Absolute Essential for Suc-
cessful Application and Interpretation of QSPR Models. QSAR
Comb. Sci. 2003, 22, 69-77.
4
1
controls.
(
MCF-7 Cytotoxicity Studies. MCF-7 cells (human breast
cancer cells) were maintained in asynchronous logarithmic
growth at 37 °C in phenol-red free Iscoves modified Dulbecco’s
medium with L-glutamine supplemented with 10% (v/v) FBS.
The population doubling time was 24-36 h. Every 48 h, the
old media was replaced by fresh media. All stock solutions and
(

7
242 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 23
Selassie et al.
(
16) Bordwell, F. G.; Zhang, X.-M.; Satish, A. V.; Cheng, J.-P.
Assessment of the Importance of Changes in Ground-State
Energies on the Bond Dissociation Enthalpies of the O-H Bonds
in Phenols and the S-H Bonds in Thiophenols. J. Am. Chem.
Soc. 1994, 116, 6605-6610.
Toxicity of Various Functional Groups. Chem. Rev. 2002, 102,
2585-2605.
(31) Hansch, C.; Gao, H. Comparative QSAR: Radical Reactions of
Benzene Derivatives in Chemistry and Biology. Chem. Rev. 1997,
97, 2995-3059.
(
17) Beck, W. T.; Cirtain, M. C. Continued Expression of Vinca
Alkaloid Resistance by CCRF-CEM Cells after Treatment with
Tunicamycin or Pronase. Cancer Res. 1982, 42, 184-189.
18) Jia, L.; Allen, P. D.; Macey, M. G.; Grahn, M. F.; Newland, A.
C.; Kelsey, S. M. Mitochondrial Electron Transport Chain
Activity but Not ATP Synthesis Is Required for Drug-Induced
Apoptosis in Human Leukemia Cells: A Possible Novel Mech-
anism of Regulating Drug Resistance. Br. J. Haematol. 1997,
(32) Beckett, A. H.; Rowland, M. Urinary Excretion Kinetics of
Amphetamine in Man. J. Pharm. Pharmacol. 1965, 17, 628-
639.
(33) Wright, J. S.; Carpenter, D. J.; McKay, D. J.; Ingold, K. U.
Theoretical Calculation of Substituent Effects on the O-H Bond
Strength of Phenolic Antioxidants Related to Vitamin E. J. Am.
Chem. Soc. 1997, 119, 4245-4252.
(34) Bosque, R.; Sales, J. A QSPR Study of O-H Bond Dissociation
Energy in Phenols. J. Chem. Inf. Comput. Sci. 2003, 43, 637-
642.
(35) Thompson, D. C.; Thompson, J. A.; Sugumaran, M.; Modeus, P.
Biological and Toxicological Consequences of Quinone Methide
Formaiton. Chem.-Biol. Interact. 1993, 86, 129-162.
(36) Bolton, J. L.; Valerio, J. L. G.; Thompson, J. A. The Enzymatic
Fomation and Chemical Reactivity of Quinone Methides Cor-
relate with Alkylphenol-Induced Toxicity in Rat Hepatocytes.
Chem. Res. Toxicol. 1992, 5, 816-822.
(37) Selassie, C. D.; Kapur, S. Unpublished results.
(38) Selassie, C. D.; Verma, R. P.; Kapur, S.; Shusterman, A. J.;
Hansch, C. QSAR for the Cytotoxicity of 2-Alkyl or 2,6-Dialkyl-
4-X-phenols: The Nature of the Radical Reaction. J. Chem. Soc.,
Perkin Trans. 2 2002, 1112-1117.
(39) Ettel, V.; Hebky, J. Derivatives of pyrone, 4(1H)-pyridone, and
pyridine. I. Preparation of Some 1-phenyl-4(1H)-pyridones.
Collect. Czech. Chem. Commun. 1950, 15, 639-652.
(40) Bessems, J. G.; Gaisser, H. D.; Te Koppele, J. M.; Van Bennekom,
W. P.; Commandeur, J. N.; Vermeulen, N. P. 3,5-Disubstituted
Analogues of Paracetamol. Synthesis, Analgesic Activity and
Cytotoxicity. Chem.-Biol. Interact. 1995, 98, 237-50.
(41) Jones, L. J.; Gray, M.; Yue, S. T.; Haugland, R. P.; Singer, V. L.
Sensitive Determination of Cell Number Using the CyQUANT
Cell Proliferation Assay. J. Immunol. Methods 2001, 254, 85-
98.
(
98, 686-698.
(
19) Kroemer, G. Mitochondrial Control of Apoptosis: An Introduc-
tion. Biochem. Biophys. Res. Commun. 2003, 304, 433-435.
20) Mat e´ s, J. M.; S a´ nchez-Jim e´ nez, F. M. Role of Reactive Oxygen
Species in Apoptosis: Implications for Cancer Therapy. Int. J.
Biochem. Cell Biol. 2000, 32, 157-170.
(
(
21) Wang, X. The Expanding Role of Mitochondria in Apoptosis.
Genes Dev. 2002, 15, 2922-2933.
(
22) Yu, R.; Mandlekar, S.; Kong, A. N. T. Molecular Mechanisms of
Butylated Hydroxy-Anisole-Induced Toxicity: Induction of Apo-
ptosis through Direct Release of Cytochrome c. Mol. Pharmacol.
2
000, 58, 431-437.
(
(
23) Selassie, C.; Kapur, S. Unpublished results.
24) Saito, M.; Sakagami, H.; Fujisawa, S. Cytotoxicity and Apoptosis
Induction by Butylated Hydroxyanisole (BHA) and Butylated
Hydroxytoluene (BHT). Anticancer Res. 2003, 23, 4693-4701.
25) Okubo, T.; Yokoyama, Y.; Kano, K.; Kano, I. Molecular Mech-
anism of Cell Death Induced by the Antioxidant tert-Butylhy-
droxyanisole in Human Monocytic Leukemia U937 Cells. Biol.
Pharm. Bull. 2004, 27, 295-302.
(
(
26) Borges dos Santos, R. M.; Martinho Simoes, J. A. Energetics of
the O-H Bond in Phenol and Substituted Phenols: A Critical
Evaluation of Literature Data. J. Phys. Chem. Ref. Data 1998,
27, 707-739.
(
27) Mulder, P.; Saastad, O. W.; Griller, D. Oxygen-Hydrogen Bond
Dissociation Energies in Para-Substituted Phenols. J. Am.
Chem. Soc. 1988, 110, 4090-4092.
(42) Etzenhouser, B.; Hansch, C.; Kapur, S.; Selassie, C. D. Mecha-
nism of Toxicity of Caffeic and Dihydrocaffeic Acids. Bioorg. Med.
Chem. 2001, 9, 199-209.
(43) Hansch, C.; Leo, A. Exploring QSAR: Fundamentals and
Applications in Chemistry and Biology; American Chemical
Society: Washington, DC, 1995.
(44) Selassie, C. The History Of Quantitative Structure Activity
Relationships. In Berger’s Medicinal Chemistry Discovery, 6th
ed.; Abraham, D. J., Ed.; John Wiley and Sons: New York, 2003;
Vol. 1, pp 1-48.
(
28) Jonsson, M.; Lind, J.; Eriksen, T. E.; Merenyi, G. Oxygen-
Hydrogen Bond Strengths and One-Electron Reduction Poten-
tials of Multisubstituted Phenols and Phenoxy Radicals. Pre-
dictions Using Free Energy Relationships. J. Chem. Soc., Perkin
Trans. 2 1993, 1567-1573.
(
29) de Heer, M. I.; Korth, H.-G.; Mulder, P. Poly Methoxy Phenols
in Solution: O-H Bond Dissociation Enthalpies, Structures, and
Hydrogen Bonding. J. Org. Chem. 1999, 64, 6969-6975.
30) Selassie, C. D.; Garg, R.; Kapur, S.; Kurup, A.; Verma, R. P.;
Mekapati, S. B.; Hansch, C. Comparative QSAR and the Radical
(
JM050567W