Involvement of semiquinone radicals in the in vitro cytotoxicity of cigarette mainstream smoke

Article abstract of DOI:10.1021/tx060162u

Free radicals in cigarette smoke have attracted a great deal of attention because they are hypothesized to be responsible in part for several of the pathologies related to smoking. Hydroquinone, catechol, and their methyl-substituted derivatives are abundant in the particulate phase of cigarette smoke, and they are known precursors of semiquinone radicals. In this study, the in vitro cytotoxicity of these dihydroxybenzenes was determined using the neutral red uptake (NRU) assay, and their radical-forming capacity was determined by electron paramagnetic resonance (EPR). All of the dihydroxybenzenes studied were found to generate appreciable amounts of semiquinone radicals when dissolved in the cell culture medium employed in the NRU assay. Hydroquinone exhibited by far the highest capacity to form semiquinone radicals at physiological pH, even though it is not the most cytotoxic dihydroxybenzene. Methyl-substituted dihydroxybenzenes were found to be more cytotoxic than either hydroquinone or catechol. The formation of semiquinone radicals via auto-oxidation of the dihydroxybenzenes was found to be dependent on the reduction potential of the corresponding quinone/semiquinone radical redox couple. The capacity to generate semiquinone radicals was found to be insufficient to explain the variance in the cytotoxicity among the dihydroxybenzenes in our study; consequently, other mechanisms of toxicity must also be involved. The observed interactions between 2,6-dimethylhydroquinone and hydroquinone in the cytotoxicity assay and EPR analysis suggest that care needs to be taken when the bioactivity of cigarette smoke constituents is evaluated, i.e., the effect of the cigarette smoke complex matrix on the activity of the single constituent studied must be taken into consideration.

Full text of DOI:10.1021/tx060162u

1602
Chem. Res. Toxicol. 2006, 19, 1602-1610
Involvement of Semiquinone Radicals in the in Vitro Cytotoxicity of
Cigarette Mainstream Smoke
Salem Chouchane,*,† Jan B. Wooten,† Franz J. Tewes,‡ Arno Wittig,‡ Boris P. Mu¨ller,§
Detlef Veltel,‡ and Joerg Diekmann‡
Philip Morris USA Research Center, 4201 Commerce Road, Richmond, Virginia 23234,
Philip Morris Research Laboratories GmbH, Fuggerstrasse 3, D-51149 Cologne, Germany,
and 58 Chapman Way, St. Neots, PE19 2HD U.K.
ReceiVed July 17, 2006
Free radicals in cigarette smoke have attracted a great deal of attention because they are hypothesized
to be responsible in part for several of the pathologies related to smoking. Hydroquinone, catechol, and
their methyl-substituted derivatives are abundant in the particulate phase of cigarette smoke, and they
are known precursors of semiquinone radicals. In this study, the in vitro cytotoxicity of these
dihydroxybenzenes was determined using the neutral red uptake (NRU) assay, and their radical-forming
capacity was determined by electron paramagnetic resonance (EPR). All of the dihydroxybenzenes studied
were found to generate appreciable amounts of semiquinone radicals when dissolved in the cell culture
medium employed in the NRU assay. Hydroquinone exhibited by far the highest capacity to form
semiquinone radicals at physiological pH, even though it is not the most cytotoxic dihydroxybenzene.
Methyl-substituted dihydroxybenzenes were found to be more cytotoxic than either hydroquinone or
catechol. The formation of semiquinone radicals via auto-oxidation of the dihydroxybenzenes was found
to be dependent on the reduction potential of the corresponding quinone/semiquinone radical redox couple.
The capacity to generate semiquinone radicals was found to be insufficient to explain the variance in the
cytotoxicity among the dihydroxybenzenes in our study; consequently, other mechanisms of toxicity
must also be involved. The observed interactions between 2,6-dimethylhydroquinone and hydroquinone
in the cytotoxicity assay and EPR analysis suggest that care needs to be taken when the bioactivity of
cigarette smoke constituents is evaluated, i.e., the effect of the cigarette smoke complex matrix on the
activity of the single constituent studied must be taken into consideration.
all of the smoke condensate collected on the Cambridge filter
pad. “Tar” is defined as the TPM less water and nicotine. The
gas-phase radicals have been shown to be unstable and very
reactive, whereas the radicals in the TPM are much longer lived
(8, 12, 16). The chemical behaviors of the GVP and TPM
radicals of cigarette smoke are quite distinct. Experimental
evidence suggests that the gas-phase radicals form in a continu-
ous mechanism, whereby nitric oxide (NO) reacts with molec-
ular oxygen in air to form NO2, and subsequent reactions
between NO2 and unsaturated molecules in cigarette smoke, e.g.,
isoprene and butadiene, yield alkyl, peroxyl, and alkoxy radicals.
Reaction of the peroxyl radicals with NO generates additional
NO2, creating a steady-state cycle (12, 17). The TPM radicals
have been tentatively assigned to semiquinone radicals in a
polymeric matrix (8). The particulate phase also contains
constituents, e.g., dihydroxybenzenes, that can generate semi-
quinone radicals (18). Hydroquinone and catechol are well-
known examples of smoke constituents that undergo oxidation
in air to form semiquinone radicals and ultimately quinones.
Further reactions between the semiquinone radicals and mo-
lecular oxygen, either in aqueous extracts of cigarette tar (ACT)
or in living cells, lead to the creation of reactive oxygen species
(ROS) such as superoxide radical, hydrogen peroxide, and
hydroxyl radical (18).
Introduction
Smoking causes disease (1-5). Extensive research has been
conducted to explain the relationship between individual smoke
constituents and smoking-related diseases, but many complex
biological processes are involved, and no simple correlations
have been found. Numerous constituents of cigarette smoke have
been identified as potential agents of biological damage,
including tobacco-specific nitrosamines, polycyclic aromatic
hydrocarbons, phenolic compounds, and free radicals (6-9).
Free radicals have attracted much attention, because free radicals
can cause DNA damage, lipid peroxidation, and protein oxida-
tion (10, 11). Cigarette smoke contains large amounts of free
radicals (8, 12) and molecular constituents that readily produce
free radicals in aqueous media (8, 13). The involvement of free
radicals in cigarette-smoke-related pathologies is plausible, but
in-depth studies are needed to delineate their precise effects.
Free radicals are found in both the gas/vapor phase (GVP)
and the particulate phase of the cigarette smoke aerosol. To
investigate the GVP and particulate-phase radicals indepen-
dently, the particulate fraction of cigarette smoke is usually
separated from the gas/vapor-phase constituents by passing the
cigarette smoke through a fiberglass filter called a Cambridge
pad (14, 15). Thus, the total particulate matter (TPM) includes
* Corresponding author. E-mail: salem.chouchane@pmusa.com.
Phone: 804-274-3937. Fax: 804-274-2576.
† Philip Morris USA Research Center.
Cytotoxicity is regarded as a potential step in several chronic
disease processes, including carcinogenesis and emphysema
(19). This might be especially true for cigarette smoke, where
the promoting events might be due to the cytotoxic activity.
‡ Philip Morris Research Laboratories GmbH.
§ 58 Chapman Way.
10.1021/tx060162u CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/01/2006

Cytotoxicity of Phenolics Found in Cigarette Smoke
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1603
The ultimate cytotoxic constituents of cigarette smoke and their
mechanisms of action are poorly understood. Thus far, most
research has focused either on the chemistry of cigarette smoke
or on the in vitro and in vivo effects of cigarette smoke on
biological systems. The lack of a bridge between these two
approaches has made it difficult to assess the relationship
between cigarette smoke constituents and their deleterious
effects in biological systems. It is known, for example, that
cigarette smoke contains a substantial amount of dihydroxy-
benzenes in the TPM (20-22). In vitro assays have demon-
strated that ACT, which contains significant amounts of
hydroquinone, catechol, and other mono- and dihydroxyben-
zenes, can damage DNA (23). However, there is no direct
evidence to suggest that pure hydroquinone or catechol can
induce the same level of free-radical formation and DNA
damage as any isolated fraction of ACT.
cigarette tobacco types (31) and is recommended by regulatory
authorities (32).
The TPM from the smoke of 2R4F research cigarettes was
collected and analyzed to determine the relative yield of the
dihydroxybenzenes. Electron paramagnetic resonance (EPR) was
used to quantify the amount of semiquinone radicals formed in
DMEM (Dulbecco’s Modified Eagle’s Medium), the growth
medium employed in the NRU assay, both at equimolar
concentrations and at concentrations proportional to their yield
in the TPM of the 2R4F cigarettes.
Experimental Procedures
Materials. Hydroquinone (HQ), 2-methylhydroquinone (2MHQ),
2,3-dimethylhydroquinone (23DMHQ), 2,6-dimethylhydroquinone
(26DMHQ), trimethylhydroquinone (TMHQ), catechol (CAT),
3-methylcatechol (3MCAT), 4-methylcatechol (4MCAT), and 4-hy-
droxy-2,2,6,6-tetramethylpiperidinooxy (TEMPO) were purchased
from Acros Organics (Morris Plains, NJ). Dulbecco’s Modified
Eagle’s Medium (DMEM) was obtained from Invitrogen Corpora-
tion (Grand Island, NY). All other chemicals were reagent grade.
Cigarette Smoking Procedure. 2R4F cigarettes (low-tar re-
search cigarettes from the University of Kentucky and Health
Research Institute) were smoked on a 20-port automatic smoking
machine under FTC smoking conditions (35-mL puff volume, 2-s
duration, 1 puff/min, with the ventilation holes of the cigarettes
left unblocked). For analysis of phenolic constituents, mainstream
smoke from 10 cigarettes was trapped on Cambridge fiberglass
filters. Four replicate smoke samples were generated per cigarette.
TPM Constituent Analysis. The phenolic constituents in the
TPM of cigarette smoke were analyzed by GC/MS. Briefly, the
TPM collected on a Cambridge pad was extracted with 10 mL of
dichloromethane/methanol (80:20), shaken for several minutes, and
centrifuged, and an aliquot of the extract was derivatized with an
excess of silylating agent (BSTFA) containing 10% TMCS. The
analysis was performed by GC/MS on an Agilent 6890 GC
equipped with an Agilent 5973 quadrupole MSD analyzer using
selected ion mode for qualitative and quantitative analysis.
Cytotoxicity Assay. Cytotoxicity was determined using the
neutral red uptake assay as described by Borenfreund and co-
workers (26, 27). Following standard operating procedures (33, 34),
3T3 cells (BALB/c mouse embryo cell line) obtained from ATCC
(American Type Culture Collection) were gown in Dulbecco’s
Modified Eagle’s Medium (DMEM) with 5% fetal bovine serum
and were exposed for 24 h to smoke fractions, or pure constituents
were added to the culture medium. For each sample or mixture,
four replicate 96-well microtiter plates were used, each with eight
test substance concentrations. Each test substance concentration was
replicated six times per microtiter plate. After exposure, the medium
was replaced by fresh medium containing neutral red dye (25 µg/
mL). Following an incubation period of 3 h, the neutral red taken
up by viable cells was determined spectrophotometrically. The
amount of neutral red taken up is directly proportional to the number
of viable cells. The cytotoxic response to the smoke was character-
ized as the EC50 value, i.e., the effective concentration that decreased
the number of viable cells by 50% relative to the solvent control.
The EC50 value is inversely related to the cytotoxicity and was
determined by nonlinear least-squares fit of the concentration-
response values to the logistic function
Even though hydroquinone is the most abundant dihydroxy-
benzene in cigarette smoke, Blakely et al. (24) showed that the
yield of particulate-phase radicals is not correlated with the yield
of hydroquinone. For a series of cigarettes containing different
tobacco blends with variable yields of hydroquinone in their
smoke TPM, the amount of free radicals in the TPM remained
unchanged. In this case, methylene chloride was used to extract
the TPM from a Cambridge pad, was solvent evaporated, and
the residue was redissolved in benzene for EPR measurements.
Under these experimental conditions, hydroquinone is less likely
to undergo auto-oxidation, i.e., oxidation due to exposure to
air, which is less favorable in organic solvents. Nevertheless,
significant levels of TPM radicals were observed, which calls
into question the exact nature of the radicals that could be
involved in the biological damage.
Neither the measurements of Pryor et al. (23) nor those of
Blakley et al. (24) are representative of radical formation under
physiological conditions or in biological systems. The dihy-
droxybenzenes can generate semiquinone radicals, quinones, and
reactive oxygen species (ROS) in oxygenated physiological
media in vitro, for example, in the growth media used in
cytotoxicity assays, and in vivo in the epithelial lining fluid of
the lungs of smokers. The cytotoxicity of quinols and quinones
has been suggested to be due to the concerted action of several
processes that include redox cycling; alteration of thiols balance
through oxidation or arylation; inhibition of cellular functions;
alteration of Ca2+ homeostasis; and covalent binding to nucleic
acids, proteins, and lipids (25). It is difficult to extrapolate in
vitro data analysis to an in vivo system, especially if the in
vitro experiments were not conducted under conditions close
to physiological conditions. Moreover, few data are available
on the cytotoxicity of pure dihydroxybenzenes to compare with
the toxicity of a complex mixture containing dihydroxybenzenes,
such as cigarette smoke TPM.
In this work, we used model systems to evaluate whether
the cytotoxicity of several pure dihydroxybenzenes found in the
TPM of cigarette smoke is related to their capacity to generate
semiquinone radicals under physiological conditions. The cy-
totoxicity of pure dihydroxybenzenes was determined using the
neutral red uptake (NRU) assay, developed by Borenfreund and
co-workers (26, 27). This assay has been widely used by the
chemical and pharmaceutical industries as an in vitro screening
method to determine the basal cytotoxicity of compounds (28).
The assay is a well-established, reproducible, standardized short-
term assay that responds to cytotoxic compounds in a dynamic
range of 5 orders of magnitude (29). The assay is known to be
responsive to both the particulate phase and the gas phase of
cigarette smoke (30). It can also discriminate between different
a
y )
c
x
1 +
( )
b
where x is the concentration, y is the absorbance relative to the
untreated control culture, a is a maximum value, and c is the slope
at the inflection point. The EC50 value is represented by the
regression parameter b. The experimental design for testing
combinations of compounds was a ray design, as described
elsewhere (35), using the individual compounds and three mixtures

1604 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Chouchane et al.
Figure 1. Chemical structures of the dihydroxybenzenes used in this study.
Table 1. Yields of Dihydroxybenzenes in TPM and Their Specific
Cytotoxicities Expressed as EC50 in the Neutral Red Uptake Assay
EC50
yield in TPM
(µg/cig)
dihydroxybenzenes
(mM)a
HQ
0.021 ( 0.004
0.011 ( 0.001
0.015 ( 0.005
0.016 ( 0.002
0.026
0.33
0.036 ( 0.005
0.052 ( 0.013
34.4 ( 2.0a
4.02 ( 0.03b
1.37 ( 0.07b
0.49 ( 0.04b
1.83 ( 0.03b
45.3 ( 0.5c
5.3
2MHQ
23DMHQ
26DMHQ
TMHQ
CAT
3MCAT
4MCAT
total
4.4
97.1
a Mean ( standard deviation, N ) 3. b Mean ( standard deviation, N )
4. c Mean ( standard deviation, N ) 10.
with constant combination ratios and serial dilution of any
combination. For the estimated response surface modeling, Sig-
maPlot 8.0 was used with smoothed data.
Figure 2. Cytotoxicities of hydroquinone and catechol. Values of the
dose-response curves represent the mean and standard deviation of
four microtiter plates.
EPR Analysis. Fresh solutions of dihydroxybenzenes were
prepared by dissolving the pure substances in 5 mL of culture
medium DMEM, 50 mM pH 7.4 phosphate buffer solution, or other
appropriate solution at equimolar concentrations of 1 mM or at
concentrations proportional to the yield of the dihydroxybenzenes
in TPM. The samples were immediately injected into an AquaX
EPR cell (Bruker Biospin Corp., Billerica, MA) mounted in a high-
sensitivity EPR cavity. The spectra were recorded at room tem-
perature using a Bruker EMX EPR instrument at X-band. The
following EPR parameters were used: microwave power, 3.98 mW;
modulation amplitude, 0.2 G; time constant, 20.48 ms; conversion
time, 20.48 ms; number of scans, 20. The concentration of free
radicals was determined by double integrating the EPR signal using
WinEPR software and comparing with known concentrations of
TEMPOL free radical. EPR simulation was used to determine the
hyperfine splitting constants and confirm the identity of the radical.
The solutions of dihydroxybenzenes were also stored in open-
air tubes for 1 or 24 h prior to analysis by EPR to study the effect
of aging on semiquinone radical formation. In addition, the
interaction of HQ with other dihydroxybenzenes was also studied.
Briefly, HQ was mixed with one constituent, and the EPR spectrum
was acquired as described above. All experiments were carried out
in triplicate, and the results are presented as the mean ( standard
deviation.
Figure 3. Estimated dose-response surface of hydroquinone, 2,6-
dimethylhydroquinone, and three combinations using smoothed data
for modeling. The vertical axis represents the relative number of viable
cells. The nonlinear contour at 50% represents the EC50 values given
in the isobologram in Figure 4.
Results
The TPM of cigarette smoke is a highly complex mixture
containing many chemical classes of constituents, including
saturated and unsaturated hydrocarbons, alcohols, aldehydes,
ketones, carboxylic acids, esters, phenols, nitriles, terpenoids,
and alkaloids (14, 15). More than 380 phenolic constituents have
been reported to be formed in the main stream of cigarette smoke
(7). Unsubstituted and methyl-substituted dihydroxybenzenes
are among the most abundant constituents in TPM. The
constituents examined in our study are shown in Figure 1.
TPM Constituent Analysis and Cytotoxicity. As reported
in Table 1, CAT and HQ exhibited the highest yields in the
TPM of mainstream cigarette smoke, whereas the methyl-
substituted HQ and CAT exhibited lower yields. The para-
dihydroxybenzenes were more cytotoxic than the ortho-

Cytotoxicity of Phenolics Found in Cigarette Smoke
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1605
Figure 4. Isobolographic characterization of interaction of hydroquinone with (A) 2,6-dimethylhydroquinone and (B) catechol on cytotoxicity.
Data represent EC50 values (95% confidence limits. The three tested mixtures for A consisted of 12/1, 4/1, and 1/1 mixtures of HQ/26DMHQ;
those for B consisted of 0.5/1, 0.1/1, and 0.03/1 mixtures of HQ/CAT. The dashed lines indicate the theoretical lines of additivity, whereas the
dotted lines indicate the used fixed-ratio rays of the mixtures.
dihydroxybenzenes, and the methyl-substituted dihydroxybenzenes
were more cytotoxic than their parent compounds, with the
exception of TMHQ. For example, the cytotoxicity of the para-
dihydroxybenzenes was in the order 2MHQ > 23DMHQ ≈
26DMHQ > HQ > TMHQ (with a factor of 2.4 between the
lowest and the highest EC50 values), and for the ortho-
dihydroxybenzenes, the order was 3MCAT > 4MCAT > CAT
(with a range of a factor of 6.4). Most of the dihydroxybenzenes
examined showed similar sigmoidal dose-response curves, as
shown for HQ in Figure 2. The exception was CAT, for which
the dose-response curve follows a high-order polynomial
equation with a long plateau of reduced number of viable cells
(i.e., cell cycle arrest during the 24 h of incubation compared
to the untreated control) in the middle section (Figure 2). The
difference in dose-response curves between catechol and
hydroquinone suggests that these two dihydroxybenzenes have
different mechanisms of cytoxicity.
To determine whether interactions between the dihydroxy-
benzenes in the TPM affect their cytotoxicity, we carried out
experiments in which HQ was mixed with another dihydroxy-
benzene and the specific cytotoxicity of the mixtures was
determined. For example, the mixtures of 26DMHQ and HQ
using constant concentration ratios induced cell viability dif-
ferently from the theoretical effect of the combination of the
pure substances, as seen in the mesh deformation in the 3D
graph (Figure 3). The contour of constant 50% viability response
in the three-dimensional dose-response curve represents the
EC50 values plotted in the two-dimensional isobologram (Figure
4A). At low concentrations of 26DMHQ and high concentrations
of HQ, the EC50 values of two of the three mixtures are higher
than the theoretical line for an additive response (dashed line
in Figure 4A). This finding indicates a lower cytotoxicity than
the sum of the cytotoxicities of the individual compounds and
therefore an antagonistic effect of the mixture of 26DMHQ and
HQ. By comparison, when HQ was mixed with CAT, the
cytotoxicities were found to be additive (the cytotoxicity of the
mixture is equal to the sum of the cytotoxicities of the individual
compounds), and therefore no interaction was observed, as
shown in Figure 4B.
Figure 5. EPR spectra of semiquinone radicals formed in the presence
of 1 mM dihydroxybenzenes in DMEM. The spectra were recorded
after 1 h of incubation. Refer to the text for experimental conditions.
medium (DMEM) that is commonly employed in the NRU
cytotoxicity assay. The pure dihydroxybenzenes were dissolved
in aerated aqueous solutions of DMEM at pH 7.4 at 1 mM
concentration. The EPR spectra of the corresponding semi-
quinone radicals were observed as shown in Figure 5. Each EPR
spectrum exhibited a pattern of hyperfine couplings character-
istic of the particular semiquinone radical, which were identical
in each case to those previously reported (37). For example, 1
mM HQ in DMEM showed a quintet signal with a 1:4:6:4:1
intensity ratio and a hyperfine splitting constant for the
interaction of the impaired electron with four equivalent protons
(a4H ) 2.37 G), in accordance with the 1,4-benzosemiquinone
anion radical (36). All other dihydroxybenzenes showed EPR
signals and hyperfine splitting constants similar to those reported
in the literature (37).
Semiquinone Radical Detection in the Presence of Dihy-
droxybenzenes. Dihydroxybenzenes are known to undergo auto-
oxidation in solution, particularly at alkaline pH (36). To gain
insight into the mechanism of cytotoxicity of dihydroxybenzenes
and to investigate the possible involvement of free radicals, we
employed EPR spectroscopy to assay the capacity of several
dihydroxybenzenes found in cigarette smoke TPM (Figure 1)
to form semiquinone radicals through auto-oxidation in the
The pH and the concentration of the dihydroxybenzene
solutions are important factors that affect the ability of the
dihydroxybenzenes to auto-oxidize and generate semiquinone
radicals. To mimic the actual yield of TPM constituents, we
also examined pure dihydroxybenzenes at concentrations pro-
portional to their yield in TPM. These concentrations are
summarized in Table 2 along with the actual yields of the
dihydroxybenzenes in the TPM of 2R4F cigarettes. When the

1606 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Chouchane et al.
Table 2. Concentrations of the Dihydroxybenzenes Used in the EPR
Experiments
dihydroxybenzenes individually in DMEM at the concentrations
reported in Table 2, and the EPR spectra were acquired. In every
case, only the EPR signal of 1,4-benzosemiquinone anion radical
was observed, except in the presence of CAT, where a very
weak signal of 1,2-benzosemiquinone anion radical was also
observed (not shown). For example, the formation of semi-
quinone radicals from HQ in the presence of any of the methyl-
substituted 1,4-dihydroxybenzenes decreased after only 10 min,
whereas 1 or 24 h after mixing, the same amount of 1,4-
benzosemiquinone radical was found (Figure 9). On the other
hand, the amount of 1,4-benzosemiquinone radicals increased
when HQ was mixed with the methyl-substituted 1,2-dyhy-
droxybenzenes. These results suggest that the interaction of
dihydroxybenzenes with HQ and the formation of semiquinone
radicals depend on the structure of the constituent and its
concentration.
To elaborate on the interaction of HQ with other dihydroxy-
benzenes, we followed the effect of each dihydroxybenzene on
the capacity of hydroquinone to generate semiquinone radicals,
particularly 26DMHQ, which shows an interaction with HQ in
the cytotoxicity assay. When 26DMHQ (10 or 100 µM) was
added to HQ (740 µM; proportional to the TPM concentration),
only 1,4-benzosemiquinone radical was observed, and its yield
was decreased at 10 min compared to HQ alone and then
appeared to be same within 1 and 24 h, as shown in Figure 10.
However, at a higher concentration of 26DMH (500 µM), only
the 2,6-dimethyl-1,4-benzosemiquinone radical was observed
(Figure 11). Within 24 h, 1,4-benzosemiquinone appeared again
(Figure 11), and its yield was almost 2 times higher than in the
case of HQ alone (Figure 10). The interaction of CAT with
HQ was different from that of 26DMHQ. When CAT was added
to HQ at different concentrations, only 1,4-benzosemiqunone
radical was observed, and no significant effect on the yield of
the radical was observed, except when the concentration of CAT
was higher than that of HQ, where an increase in the semi-
quinone radical yield was observed (Figure 12).
yield in TPM
concentration in
dihydroxybenzene
(µg/cig)
EPR assay (µM)
HQ
34.4
4.02
1.37
0.49
1.83
45.3
5.3
743
67
27
10
44
819
85
71
2MHQ
23DMHQ
26DMHQ
TMHQ
CAT
3MCAT
4MCAT
4.4
Table 3. One-Electron Reduction Potentials (mV, 25 °C, pH 7.0) of
the Redox Couples (Q/Q-•
)
E(Q/Q-•)a at pH 7
(mV)
HQ
78
23
-74
-80
-165
2MHQ
23DMHQ
26DMHQ
TMHQ
a Taken from ref 38.
dihydroxybenzenes were dissolved in either DMEM or phos-
phate buffer, significant amounts of semiquinone radicals formed
in solutions of HQ, 2MHQ, and CAT, but none of the other
compounds (Figure 6). The overall yield of semiquinone radicals
was slightly higher in DMEM (Figure 6B) than in phosphate
buffer (Figure 6A). Semiquinone radical formation from 2MHQ
was observed in DMEM, but not in phosphate buffer. HQ
produced the highest yield of semiquinone radicals, whereas
CAT did not generate a significant amount of radicals, even
with a concentration higher than that of HQ. When dissolved
in water, DMSO, or ethanol at TPM concentration, only HQ
generated semiquinone radicals in significant amounts (not
shown).
Because the concentrations used above were different for each
dihydroxybenzene and because only HQ, CAT, and 2MHQ
generated measurable amounts of their corresponding semi-
quinone radicals, we also examined the radical formation for
solutions of the dihydroxybenzenes at equimolar concentrations
(1 mM). As shown in Figure 7, semiquinone radicals were
readily detected for all eight dihydroxybenzenes; however, the
increase in the amount of radicals did not increase in proportion
to the concentration of the parent dihydroxybenzene. In addition,
it appears that semiquinone formation is favored when the
hydroxyl groups are in the para position (e.g., HQ) compared
to the ortho position (e.g., CAT).
The reduction potential is known to be an essential parameter
that governs the auto-oxidation of dihydroxybenzene com-
pounds. In Table 3, we summarize the literature data on one-
electron reduction potentials of quinone/semiquinone radical
redox couples (Q/Q-•) for some of the dihydroxybenzenes (38).
As shown in Figure 8, a linear correlation was found between
the structure of the dihydroxybenzene and its reduction potential.
In general, methyl substitution decreases the reduction potential
of the dihydroxybenzene. Because no reduction potential for
Q/Q-• at pH 7 was reported for catechol in the literature, we
compared the reduction potential of QH-/Q2- redox couples
for catechol and hydroquinone, 459 and 530 mV, respectively
(39). Judging from these values, Q2- of HQ appears to be more
reducing and should generate semiquinone radicals more easily
than Q2- of CAT.
Discussion
The dihydroxybenzenes in the TPM of cigarette smoke are
an important class of constituents that can act either as pro-
oxidants or as anti-oxidants. Consequently, it is not surprising
that the dihydroxybenzenes exhibit significant cytotoxicity. It
has long been known that HQ can be oxidized to benzoquinone,
a potent hemeatotoxic, genotoxic, and carcinogenic agent (40,
41). Moreover, xenobiotic HQ is known to cause the accumula-
tion of free radicals in the intracellular environment, which can
lead to chromosomal aberration and cell damage through
oxidative stress (42-44). HQ and CAT in cigarette smoke TPM
have both been shown to block lymphocyte proliferation and
inhibit T cell activation in the lung by inhibition of ribonucle-
otide reductase (binding to tyrosyl radical), the rate-limiting
enzyme in DNA synthesis (45, 46). However, the observed
differences between the dose-response curves of catechol and
hydroquinone in the present study suggest that the underlying
cytotoxic processes are different for these two substances. The
shape of the dose-response curve for catechol, with its long
plateau at intermediate concentrations (0.1-0.3 mM), suggests
that CAT is merely cytostatic up to concentrations of 0.3 mM
but cytotoxic at higher concentrations.
Our data show that the methyl-substituted dihydroxybenzenes
are more cytotoxic than the non-methylated parent compounds,
a result that is consistent with previous studies. Moridani et al.
(47), for example, reported that substituted phenols are more
cytotoxic when they are lipophilic and have lower redox
Additional experiments were performed to study the interac-
tion of HQ with other dihydroxybenzenes on semiquinone
radical formation. HQ was mixed with each of the other

Cytotoxicity of Phenolics Found in Cigarette Smoke
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1607
Figure 6. Yield of semiquinone radicals formed during the auto-oxidation of dihydroxybenzenes in (A) 50 mM pH 7.4 potassium phosphate and
(B) DMEM. The concentrations of the dihydroxybenzenes used were proportional to their yields in the TPM (see Table 2).
Figure 7. Yield of semiquinone radicals formed during the auto-
oxidation of 1 mM dihydroxybenzenes in DMEM.
Figure 9. Effect of interactions of hydroquinone with other dihy-
droxybenzenes in DMEM. The concentrations used were those reported
in Table 2.
Figure 8. Plot of the yield of semiquinone radicals formed when 1
mM of dihydroxybenzenes was dissolved in DMEM versus the
reduction potential of Q/Q-•
.
Figure 10. Semiquinone radical yield in the presence of HQ and
26DMHQ at different concentrations in DMEM. The y axis represents
the total yield of semiquinone radical in solution.
potentials, whereas substituted catechols with higher lipophilicity
and distribution coefficients, lower degree of ionization, and
higher pKa are more toxic than other catechols (48). In a study
of quantitative structure-activity relationships (QSARs) of
phenols, Smith et al. (7) showed that substituted phenols with
electron-releasing groups have a greater potential to produce
toxic phenoxyl free radicals. However, in our study, whereas
mono- and dimethyl substitution had this effect, trimethyl
substitution did not.
In certain cases, combinations of dihydroxybenzenes can
exhibit cooperative cytotoxic effects, as we have shown for
certain concentrations of HQ and 26DMHQ, thus affecting the
specific cytotoxicity of the mixture. The apparent cytotoxicity
of HQ was reduced (less than additive) by 26DMHQ at low
concentrations relative to the EC50 value of 26DMHQ. In
contrast, higher concentrations of 26DMHQ resulted in a more
additive cytotoxicity. This indicates that the two compounds
display conflicting actions, probably depending on their relative
redox potential and their concentration by as-yet-unidentified
mechanisms. However, CAT induced only an additive effect
when combined with hydroquinone. The observation that
dihydroxybenzenes can interact in the cytotoxicity assay raises
the question of whether literature data reporting the biological
effects of pure HQ or CAT are representative of their behavior
in a complex matrix, as in cigarette smoke TPM.

1608 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Chouchane et al.
with the relative position of hydroxyl groups in the dihydroxy-
benzenes. As we have shown, HQ is more toxic and generates
more semiquinone radicals than CAT. It was previously shown
that the relative ease of auto-oxidation of dihydroxybenzenes
is in the order para > ortho > meta (39).
Multiple mechanisms might be involved in the toxicity of
dihydroxybenzenes. For example, a study of the structure-
toxicity relationships for phenols in isolated rat hepatocytes
showed that one or a combination of mechanisms, e.g.,
mitochondrial uncoupling, phenoxyl radicals, or phenol me-
tabolism to quinine, can contribute to the cytotoxicity of phenols
toward hepatocytes (47). Two plausible mechanisms by which
dihydroxybenzenes can damage cellular macromolecules are (1)
the covalent binding of the reactive metabolites (semiquinone
radicals and quinones) to essential macromolecules such as
protein and DNA and (2) oxidative stress resulting from the
formation of ROS via the redox cycling of quinones through
semiquinone radical intermediates. Formation of superoxide
radicals, hydrogen peroxide, and hydroxyl radicals in ACT has
been demonstrated (23), suggesting that dihydroxybenzenes are
involved in smoking-related oxidative stress.
Figure 11. EPR spectra of semiquinone radicals when 500 µM of
26DMHQ was mixed with HQ as function of time.
Because of the large number of constituents found in cigarette
smoke TPM, chemical interactions between the constituents or
cooperative toxicological effects might affect the cytotoxicity
of the TPM. In the NRU cytotoxicity assay, an antagonistic
effect was observed when hydroquinone was mixed with
26DMHQ. Similarly, EPR data showed that some dihydroxy-
benzenes can interact with hydroquinone and alter the yield of
semiquinone radicals, whereas other combinations had no effect.
A simple relationship, however, was not found between the
radical yield and the cytotoxicity of the mixture. There is a
paucity of data to indicate whether the critical interactions occur
in the extracellular cell growth medium or within the cells. Other
factors might influence cytotoxicity and should also be taken
into account; for example, the presence of transition metals in
the TPM. Iron and copper have been shown to enhance the
oxidation of hydroquinone in a concentration-dependent manner
(50-52). It has been shown that, in the presence of NADH,
CAT induced more Cu2+-mediated DNA damage than HQ,
whereas the reverse relation was observed in the absence of
NADH (53). The peroxidase-dependent conversion of hydro-
quinone to benzoquinone has been shown to be stimulated by
phenolic constituents such as catechol, resorcinol, and cresol
(54).
Figure 12. Semiquinone radical yield in the presence of HQ and CAT
at different concentrations in DMEM on semiquinone radical yield.
The y axis represents the total yield of semiquinone radical in the
solution.
It is well-known that that semiquinone radical formation from
HQ and CAT is favored in alkaline solution (36). The results
of Pedersen (36), as well as our own, show that semiquinone
radicals can also be generated at physiological pH. Moreover,
we have shown that the dihydroxybenzenes undergo auto-
oxidation leading to the formation of semiquinone radicals in
cell culture media, such as DMEM.
The reactivity, thermodynamic, and redox properties of
quinones, semiquinones, and the parent quinols are of major
importance in understanding their biological activity. As we have
shown, 2MHQ, 23DMHQ, and 26DMHQ are more cytotoxic
than HQ, but at the same concentration, they generate smaller
amounts of semiquinone radicals. This suggests that the cyto-
toxicity of the methyl-substituted dihydroxybenzenes is not
solely dependent on radical formation and that other mechanisms
should be considered. Smith et al. (7) have suggested that the
electron-donating character of methyl groups stabilizes the
semiquinone radical, thereby prolonging its lifetime and enhanc-
ing its potential for biological damage. Thus, the kinetics of
semiquinone radical formation can be a determining factor in
the cytotoxicity of dihydroxybenzenes. However, for dihydroxy-
benzenes lacking an electron-donating group, the cytotoxicity
can be dominated by other factors, e.g., hydrophobicity (7).
A correlation between the cytotoxicity of quinones and their
reduction potentials has been previously reported (49). It was
observed, for example, that the toxicity of natural hydroxyan-
thraquinones increases at pH 7 with an increase in reduction
potential, pointing to an oxidative stress mechanism (49). The
cytotoxicity and the capacity to form semiquinone radicals vary
Summary and Conclusions. The dihydroxybenzenes in our
study, constituents in the TPM of cigarette smoke, were found
to exhibit significant cytotoxicity. The methyl-substituted di-
hydroxybenzenes were shown to have higher cytotoxicity than
the unsubstituted compounds. The dihydroxybenzenes were
shown to generate semiquinone radicals in DMEM, the medium
used in the NRU cytotoxicity assay. Nevertheless, a correlation
between the abundance of semiquinone radicals formed and
cytotoxicity was not found. The observed interaction between
2,6-dimethylhydroquinone and hydroquinone in the cytotoxicity
assay and EPR analysis demonstrate that the EC50 values in
binary mixtures of the dihydroxybenzenes cannot, in general,
be assumed to be additive. Consequently, the interpretation of
the bioactivity of cigarette smoke evaluated by similar methods
should consider the possible effects of the complex TPM matrix
on the activities of the individual constituents.
Acknowledgment. We thank Dr. Geoffrey Chan from Philip
Morris USA and Dr. Wolf Reininghaus, Cologne, Germany,
for reviewing this article.

Cytotoxicity of Phenolics Found in Cigarette Smoke
Chem. Res. Toxicol., Vol. 19, No. 12, 2006 1609
(25) Brunmark, A., and Cadenas, E. (1989) Redox and addition chemistry
of quinoid compounds and its biological implications. Free Radical
Biol. Med. 7, 435-477.
(26) Borenfreund, E., and Puerner, J. A. (1985) Toxicity determined in
vitro by morphological alterations and neutral red absorption. Toxicol.
Lett. 24, 119-124.
(27) Babich, H., and Borenfreund, E. (1992) Cytotoxic and morphological
effects of phenylpropanolamine, caffeine, nicotine, and some of their
metabolites studied in vitro. Toxicol. in Vitro 6, 493-502.
(28) National Institutes of Health (2001) Report of the international
workshop on in vitro methods for assessing acute systemic toxicity,
National Institutes of Health Publication No. 01-4499, National
Toxicology Program (NTP), Research Triangle Park, NC.
(29) Borenfreund, E., Babich, H., and Martin-Alguacil, N. (1988) Com-
parisons of two in vitro cytotoxicity assayssThe neutral red (NR)
and tetrazolium MTT tests. Toxicol. in Vitro 2, 1-6.
(30) Tewes, F. J., Meisgen, T. J., Veltel, D. J., Roemer, E., and Patskan,
G. (2003) Toxicological evaluation of an electrically heated cigarette.
Part 3: Genotoxicity and cytotoxicity of mainstream smoke. J. Appl.
Toxicol 23, 341-348.
(31) Bombick, D. W., Putnam, K., and Doolittle, D. J. (1998) Comparative
cytotoxicity studies of smoke condensates from different types of
cigarettes and tobaccos. Toxicol. in Vitro 12, 241-249.
(32) Health Canada (2004) Neutral Red Uptake Assay for Mainstream
Tobacco Smoke (http://www. hc-sc. gc. ca/hl-vs/tobac-tabac/legisla-
tion/reg/indust/propose/tox_e.html), Official Method T-502, version
1.1.
(33) National Institutes of Health (2001) Standard Operating Procedure
(SOP) for the BALB/c 3T3 Neutral Red Uptake Cytotoxicity Test (A
Test for basal Cytotoxicity). In Guidance Document on Using In Vitro
Data to Estimate In ViVo Starting Doses for Acute Toxicity, National
Institutes of Health Publication No. 01-4500, Appendix C, National
Toxicology Program (NTP), Research Triangle Park, NC.
(34) INVITTOX (1990) The FRAME Modified Neutral Red Uptake
Cytotoxicity Test, INVITTOX Protocol No. 3a, INVITTOX, Notting-
ham, U.K.
(35) Michaud, J. P., Gandolfi, A. J., and Brendel, K. (1995) Methods of
assessing toxic interactions in vitro: Experimental design and data
analysis. Toxicol. Methods 5, 21-40.
(36) Pedersen, J. A. (2002) On the application of electron paramagnetic
resonance in the study of naturally occuring quinones and quinols.
Spectrochim. Acta A 58, 1257-1270.
(37) Pedersen, J. A. (1985) Handbook of EPR Spectra from Quinones and
Quinols, CRC Press, Boca Raton, FL.
(38) Wardman, P. (1989) Reduction potentials of one-elecron couples
involving free radicals in aqueous solution. J. Phys. Chem. Ref. Data
18, 1637-1755.
(39) Steenken, S., and Neta, P. (1979) Electron transfer rates and
equilibriums between substituted phenoxide ions and phenoxyl radicals.
J. Phys. Chem. 83, 1134-1137.
(40) Snyder, R., and Hedli, C. C. (1996) An overview of benzene
metabolism. EnViron. Health Perspect. 104, 1165-1171.
(41) Bolton, J. L., Trush, M. A., Penning, R. M., Dryhurst, L., and Monks,
T. J. (2000) Role of quinones in toxicology. Chem. Res. Toxicol. 13,
135-160.
(42) Rao, N. R., and Snyder, R. (1999) Oxidative modifications produced
in HL-60 cells on exposure to benzene metabolites. J. Appl. Toxicol.
15, 403-409.
References
(1) Peto, R., Boreham, J., Lopez, A. D., Thun, M., and Heath, C. (1992)
Mortality from tobacco in developed countries: Indirect estimation
from national vital statistics. Lancet 339, 1268-1278.
(2) U.S. Department of Health and Human Services (1998) Reducing the
Health Consequences of Smoking: 25 Years of Progress. A Report of
the Surgeon General, U.S. Department of Health, Center for Disease
Control, Center for Chronic Disease Prevention and Health Promotion,
Office on Smoking and Health, DHHS Publ. No. (CDC) 89-8411.
(3) Pryor, W. A. (1987) The free radical chemistry of cigarette smoke
and the inactivation of alpha-1-proteinase inhibitor. In Pulmonary
Emphysema and Proteolysis (Taylor, J. C., and Mitmtman, C., Eds.)
pp 369-392, Academic Press, New York.
(4) Janoff, A. (1985) Elastases and emphysema. Current assessment of
the protease-antiprotease hypothesis. Am. ReV. Resp. Dis. 132, 417-
433.
(5) Janoff, A., Carp, H., Laurent, P., and Raju, L. (1983) The role of
oxidative processes in emphysema. Am. ReV. Resp. Dis. 127, 31-38.
(6) Hoffmann, D., and Hecht, S. S. (1989) Advances in tobacco carcino-
genesis. In Chemical Carcinogenesis and Mutagenesis (Cooper, C.
S., and Grover, P. L., Eds.) pp 64-102, Springer-Verlag, New York.
(7) Smith, C. J., Perfetti, T. A., Morton, M. J., Rodgman, A., Garg, R.,
Selassie, C. D., and Hansch, C. (2002) The relative toxicity of
substituted phenols reported in cigarette mainstream smoke. Toxicol.
Sci. 69, 265-278.
(8) Pryor, W. A., Hales, B. J., Premovic, P. I., and Church, D. F. (1983)
The radicals in cigarette tar: Their nature and suggested physiological
implications. Science 220, 425-427.
(9) Church, D. F., and Pryor, W. A. (1985) Free-radical chemistry of
cigarette smoke and its toxicological implications. EnViron. Health
Perspect. 64, 111-126.
(10) Baskin, S. I., and Salem, H. (1997) Oxidants, Antioxidants and Free
Radicals, Taylor and Francis, Washington, DC.
(11) Halliwell, B., and Gutteridge, J. M. C. (1999) Free Radicals in Biology
and Medicine, Oxford University Press, New York.
(12) Pryor, W. A., Prier, D. G., and Church, D. F. (1983) Electron-spin
resonance study of mainstream and sidestream cigarette smoke: Nature
of the free radicals in gas-phase smoke and in cigarette tar. EnViron.
Health Perspect. 47, 345-355.
(13) Cosgrove, J. P., Borish, E. T., Church, D. F., and Pryor, W. A. (1985)
The metal-mediated formation of hydroxyl radical by aqueous extracts
of cigarette tar. Biochem. Biophys. Res. Commun. 132, 390-396.
(14) Dube, M. F. R., and Green, C. R. (1982) Methods of collection of
smoke for analytical purposes. Rec. AdV. Tob. Sci. 8, 42-102.
(15) Baker, R. R. (1999) Smoke Chemistry: Tobacco Production, Chemistry
and Technology, Blackwell Science, London.
(16) Pryor, W. A., Terauchi, K., and Davis, W. H., Jr. (1976) Electron
spin resonance (ESR) study of cigarette smoke by use of spin trapping
techniques. EnViron. Health Perspect. 16, 161-176.
(17) Pryor, W. A., Tamura M., Dooley M. M., Premovic P., Hales, B. J.,
and Church, D. F. (1983) Reactive oxy-radicals from cigarette smoke
and their physiological effects. In Oxy Radicals and Their ScaVenger
System. Cellular and Medical Aspects (Greenwald, R., and Cohen,
G., Eds.) pp 185-192, Elsevier Science Publishing, New York.
(18) Zang, L. Y., Stone, K., and Pryor, W. A. (1995) Detection of free
radicals in aqueous extracts of cigarette tar by electron spin resonance.
Free Radical Biol. Med. 19, 161-167.
(19) Butterworth, B. E., Popp, J. A., Conolly, R. B., and Goldsworthy, T.
L. (1992) Chemically induced cell proliferation in carcinogenesis. In
Mechanisms of Carcinogenesis in Risk Identification (Vaino, H.,
Magee, P. N., McGregor, D. B., and McMichael, A. J., Eds.) pp 279-
305, International Agency for Research on Cancer (IRAC), Lyon,
France.
(20) Nanni, E. J., Lovette, M. E., Hicks, R. D., Fowler, K. W., and
Borgerding, M. F. (1990) Separation and quantitation of phenolic
compounds in mainstream cigarette smoke by capillary gas chroma-
tography with mass spectrometry in selected-ion mode. J. Chromatogr.
A 505, 365-374.
(21) Risner, H., and Cash, S. L. (1990) High performance liquid chro-
matographic determination of major phenolics compounds in tobacco
smoke. J. Chromatogr. Sci. 28, 239-244.
(43) Andreoli, C., Rossi, S., Leopardi, P., and Crebelli, R. (1999) DNA
damage by hydroquinone in human white blood cells: Analysis by
alkaline single-cell gel electrophoresis. Mutat. Res. 438, 37-45.
(44) Silva, M. D. C. U., Gaspar, J., Silva, I. D., Leao, D., and Rueff, J.
(2003) Mechanisms of induction of chromosomal aberrations by
hydroquinone in V79 cells. Mutagenesis 18, 491-496.
(45) McCue, J. M., Link, K. L., Eaton, S. S., and Freed, B. M. (2000)
Exposure to cigarette tar inhibits ribonucleotide reductase and blocks
lymphocyte proliferation. J. Immunol. 165, 6771-6775.
(46) McCue, J. M., Lazis, S., Cohen, J. J., Modiano, J. F., and Freed, B.
M. (2003) Hydroquinone and catechol interfere with T cell cycle entry
and progression through the G(1) phase. Mol. Immunol. 39, 995-
1001.
(47) Moridani, M. Y., Siraki, A., and O’Brien, P. J. (2003) Quantitative
structure toxicity relationships for phenols in isolated rat hepatocytes.
Chem.-Biol. Interact. 145, 213-223.
(48) Moridani, M. Y., Siraki, A., Chevaldina, T., Scobie, H., and Brien, P.
J. (2004) Quantitative structure toxicity relationships for catechols in
isolated rat hepatocytes. Chem.-Biol. Interact. 147, 297-307.
(49) Nemeikaite-Ceniene, A., Sergediene, E., Nivinskas, H., and Cenas,
N. (2002) Cytotoxicity of natural hydroxyanthraquinones: Role of
oxidative stress. Z. Naturforsch. C 57, 822-827.
(22) Snook, M., Chortyk, O. T., and Arrendale, R. F. (1985) Isolation and
identification of the phenolic acids of tobacco smoke. Tob. Sci. 26,
25-31.
(23) Pryor, W. A., Stone, K., Zang, L. Y., and Bermudez, E. (1998)
Fractionation of aqueous cigarette tar extracts: Fractions that contain
the tar radical cause DNA damage. Chem. Res. Toxicol. 11, 441-
448.
(24) Blakley, R. L., Henry, D. D., and Smith, C. J. (2001) Lack of
correlation between cigarette mainstream smoke particulate phase
radicals and hydroquinone yield. Food Chem. Toxicol. 39, 401-406.
(50) Li, Y., and Trush, M. A. (1993) DNA damage resulting from the
oxidation of hydroquinone by copper: Role for a Cu(II)/Cu(I) redox
cycle and reactive oxygen generation. Carcinogenesis 14, 1303-1311.

1610 Chem. Res. Toxicol., Vol. 19, No. 12, 2006
Chouchane et al.
(51) Li, Y., and Trush, M. A. (1993) Oxidation of hydroquinone by
copper: Chemical mechanism and biological effects. Arch. Biochem.
Biophys. 300, 346-355.
responsible for their differential DNA-damaging ability. Chem. Res.
Toxicol. 15, 76-82.
(54) Subrahmanyam, V. V., Kolachana, P., and Smith, M. T. (1991)Me-
tabolism of hydroquinone by human myeloperoxidase: Mechanisms
of stimulation by other phenolic compounds. Arch. Biochem. Biophys.
286, 76-84.
(52) Stohs, S. J., Bagchi, D., and Bagchi, M. (1997) Toxicity of trace
elements in tobacco smoke. Inhalation Toxicol. 9, 867-890.
(53) Hirakawa, K., Oikawa, S., Hiraku, Y., Hirosawa, I., and Kawanishi,
S. (2002) Catechol and hydroquinone have different redox properties
TX060162U