Carcinogenesis vol.22 no.1 pp.171–177, 2001
Aryl hydrocarbon receptor signaling plays a significant role in
mediating benzo[a]pyrene- and cigarette smoke condensateinduced cytogenetic damage in vivo
Stephen D.Dertinger1, Daniel A.Nazarenko1,
Allen E.Silverstone2 and Thomas A.Gasiewicz1,3
1Department
of Environmental Medicine, University of Rochester School of
Medicine, Rochester, NY 14642, USA and 2Department of Microbiology
and Immunology, State University of New York, Upstate Medical
University, Syracuse, NY 13210, USA
3To
whom correspondence should be addressed
Email: [email protected]
This laboratory has previously reported data suggesting
that aryl hydrocarbon receptor (AhR) signaling may have
a net potentiating effect on the DNA damaging potential
of cigarette smoke. The experiments described in this
report extend these studies by testing whether the potent
AhR antagonist 3⬘-methoxy-4⬘-nitroflavone (3⬘M4⬘NF) can
modify the in vivo genetic toxicity of benzo[a]pyrene (B[a]P)
and the complex mixture of chemicals in cigarette smoke
condensate (CSC). Initial experiments were designed to
determine 3⬘M4⬘NF doses which can antagonize AhR in vivo
but which have little effect on constitutive cytochrome
P4501A (CYP1A) activity. These experiments took three
forms: (i) zoxazolamine paralysis tests, a functional assay
of cytochrome P450 CYP1A activity in 3⬘M4⬘NF-treated
C57Bl/6J mice; (ii) co-treatment of Ahr null allele mice with
150 mg/kg B[a]P plus a range of 3⬘M4⬘NF concentrations in
order to evaluate the potential of the flavone to interact
with non-AhR targets which may affect B[a]P toxicity; (iii)
an evaluation of the in vivo AhR antagonist activity of
3⬘M4⬘NF using transgenic mice which carry a dioxinresponsive element-regulated lacZ reporter. Once an appropriate dose range was determined, C57Bl/6J mice were
challenged with B[a]P or CSC with and without 3⬘M4⬘NF
co-treatment. Chromosome damage was measured by scoring the frequency of micronuclei in peripheral blood reticulocytes. Data presented herein suggest that 3⬘M4⬘NF can
protect mice from B[a]P-induced bone marrow cytotoxicity
and genotoxicity. Furthermore, CSC-associated genotoxicity was abolished by the flavonoid. These data add support
to our hypothesis that AhR signaling has a net potentiating
effect on the genetic toxicity and, presumably, carcinogenicity of cigarette smoke.
Introduction
Aryl hydrocarbon receptor (AhR) signaling is known to be an
important pathway by which enzymes are induced in response
to cigarette smoke. Löfroth and Rannug (1) have shown
that uncharacterized chemicals contained in cigarette smoke
condensate compete with radiolabeled 2,3,7,8-tetracholordibenzo-p-dioxin (TCDD) for receptor binding and Gebremichael
Abbreviations: AhR, aryl hydrocarbon receptor; B[a]P, benzo[a]pyrene;
CSC, cigarette smoke condensate; DMSO, dimethyl sulfoxide; DRE, dioxinresponsive element; β-gal, β-galactosidase; 3⬘M4⬘NF, 3⬘-methoxy-4⬘-nitroflavone; MN-RET, micronucleated reticulocyte; PI, propidium iodide; RET,
reticulocyte; TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin.
© Oxford University Press
et al. (2) have demonstrated that these chemicals are capable
of transforming the receptor to an active transcription factor
in vitro. Both phase I enzymes (most notably CYP1A1, 1A2
and 1B1) and phase II enzymes are regulated by the AhR and
are known to either bioactivate procarcinogens in tobacco
smoke or to sequester and detoxify reactive electrophiles
(3–5). It is therefore plausible that AhR-induced enzyme
activity has a measurable and biologically significant effect
on the carcinogenicity of cigarette smoke. Whether it mediates
a net increase or decrease in tumorigenicity is difficult to
predict. However, some indirect evidence suggests that AhR
signaling may result in a net potentiation of cigarette smoke
carcinogenicity. For instance, studies have been published
which link high aryl hydrocarbon hydroxylase inducibility (a
function of CYP1A activity) with increased cancer risk (6,7).
Furthermore, this laboratory has demonstrated that mutant
AhR cell lines and Ahr null allele mice are relatively resistant
to chromosome damage caused by cigarette smoke condensate
(CSC) compared with wild-type controls (8). Additionally, a
recent report observed that while benzo[a]pyrene (B[a]P), an
important constituent of cigarette smoke, induced tumors in
wild-type mice, no tumors were observed in Ahr null allele
animals (9). Additional evidence is supplied by a cohort of
TCDD-exposed individuals (male factory workers). Increased
total cancer deaths were observed in individuals who actively
smoked as exposure to TCDD increased. High TCDD exposure
levels were not observed to enhance total cancer deaths in
non-smokers or ex-smokers (10).
The studies outlined above suggest that induction of the
AhR gene battery may favor bioactivation of tobacco smoke
mutagens relative to detoxification processes. Even so, it is
clear that no direct evidence is available which describes the
consequence of AhR-mediated signal transduction on the
genotoxic and carcinogenic potential of cigarette smoke. For
instance, even experiments with Ahr null allele mice do
not provide formal proof that under physiologically relevant
circumstances AhR signaling significantly affects cigarette
smoke toxicity. Since knockout animals express a low basal
level of CYP1A2 activity relative to controls, reduced by
~90% (11), it is not clear whether differential sensitivity to
cigarette carcinogens are related to altered constitutive enzyme
expression profiles or the lack of AhR-dependent induction
processes. For the work described herein we utilized a chemical
AhR antagonist, which could be used to temporally block
signaling, to discriminate between these two possibilities.
Results from cell-free and cell culture studies indicate
that 3⬘-methoxy-4⬘-nitroflavone (3⬘M4⬘NF) is a potent AhR
antagonist with little or no agonist activity (12–14). Furthermore, we have demonstrated that 3⬘M4⬘NF is able to modify
the in vivo genotoxicity of B[a]P (15). However, these same
in vivo data also suggested that at the single high concentration
of antagonist tested significant non-AhR targets are affected.
A likely secondary target is inhibition of CYP1A enzyme
activity (15). Thus, these studies did not clearly determine
171
S.D.Dertinger et al.
whether the attenuation of B[a]P genotoxicity was due to
inhibition of the AhR signaling pathway or CYP1A activity.
The experiments described in the current report were performed
to test whether 3⬘M4⬘NF can attenuate B[a]P- and CSCinduced genotoxicity at concentrations which do not significantly affect basal CYP1A activity.
Materials and methods
Reagents
Absolute methanol was purchased from Fisher Scientific (Springfield, NJ).
Sodium heparin, propidium iodide (PI), RNase A, B[a]P, Triton X-100,
phenylmethylsulfonyl fluoride, leupeptin and dithiothreitol were obtained from
Sigma (St Louis, MO). 3⬘M4⬘NF was synthesized and purified (⬎98%) by
procedures previously described (16). TCDD was purchased from Cambridge
Isotopes (Cambridge, MA). Anti-CD71–FITC was purchased from BioDesign
(clone no. R17217.1.4; Kennebunk, ME). Fixed malaria-infected mouse
erythrocytes were purchased from Litron Laboratories (Rochester, NY). CSC
was generated from University of Kentucky reference cigarettes, type 1R4F,
as described previously (8). For injection, B[a]P, zoxazolamine and TCDD
were prepared in olive oil. CSC stock [100 mg/ml dimethyl sulfoxide (DMSO)]
was diluted in olive oil to provide 2.5 mg/ml for injection. In early studies
3⬘M4⬘NF was prepared in DMSO. We switched to olive oil after it became
apparent that precipitates periodically formed in the intraperitoneal cavity of
treated mice when DMSO was the vehicle.
Animals
Male wild-type C57Bl/6J mice were purchased from Jackson Laboratories
(Bar Harbor, ME). Male Ahr null allele mice (Ahr–/–, exon 1 targeted) were
bred at SUNY (Upstate Medical University, Syracuse, NY). The founder mice
of this colony were originally developed by and received from P.FernandezSalguero and F.Gonzalez (National Cancer Institute, Bethesda, MD) (17).
Transgenic mice (DRE–lacZ) used in these studies have been described
previously (18). Briefly, the DREs in this reporter are bona fide sites from
the CYP1A1 promoter. The transgenic mice were developed by microinjecting
DNA containing the DRE–lacZ construct into pronuclei of recently fertilized
C57Bl/6J⫻SJL ova. A PCR screening method was used to identify heterogzygous, transgene-positive mice (18). Note that at the time of the current
experiments the transgene-positive line had been backcrossed to the C57Bl/
6J background more than nine times. Purina Mills Rodent Chow 5001 and
water were available to mice ad libitum. For all experiments mice were
allowed to acclimatize for at least 1 week.
Dose range finding experiment: zoxazolamine paralysis time
The zoxazolamine paralysis test (19) was used to evaluate 3⬘M4⬘NF doses
for their ability to inhibit CYP1A1/2 activity in vivo. Zoxazolamine is a potent
muscle relaxant which is metabolically inactivated by CYP1A1/2 enzyme
activity (and to a lesser extent CYP2E1). In addition to non-invasively
phenotyping mice as AhR-responsive and non-responsive (20), it has also
been used as a functional test to evaluate chemicals for their ability to inhibit
CYP1A1/2 enzyme activity in vivo (21,22). Three independent experiments
were performed with male C57Bl/6J mice, 6–9 weeks old. Mice were treated
with 0, 0.2, 2, 20 or 40 mg/kg 3⬘M4⬘NF i.p. Approximately 30 min after the
3⬘M4⬘NF injections mice were treated i.p. with zoxazolamine (125–150 mg/
kg). Paralysis times were determined as the length of time required for mice
to regain their righting reflex. Data are expressed as relative zoxazolamine
paralysis times, normalized to the average paralysis time observed for the
vehicle-treated animals in the same experiment (set at 1.0). Mean fold
induction for each 3⬘M4⬘NF exposure group was compared with the vehicle
control group by unpaired, two-tailed t-tests (StatView v.4.5). P ⬍ 0.05 was
considered a significant effect.
Dose range finding experiment: Ahr null allele mice and splenic lesions
Previous experiments have indicated that a high 3⬘M4⬘NF concentration can
enhance the systemic toxicity of B[a]P (14). The spleen was found to be a
particularly sensitive target of the synergistic toxicity observed when mice
were co-treated with B[a]P and 3⬘M4⬘NF. Interestingly, Ahr null allele mice
were found to exhibit the same types of splenic lesions as wild-type animals.
This indicates that the synergistic toxicity observed is mediated by the ability
of the flavonoid to interact with a non-AhR target(s). We used this same
system to evaluate the ability of the flavonoid to affect secondary (nonAhR) targets.
For this experiment male Ahr–/– mice were treated i.p. with 150 mg/kg
B[a]P. The flavonoid 3⬘M4⬘NF was administered as a split dose in an effort
to maintain plasma levels. Specifically, 3⬘M4⬘NF was administered i.p. at 0,
0.2, 2 or 20 mg/kg/injection 4 h prior to, concurrently with and 4 h after
172
B[a]P. Forty-eight hours after the B[a]P bolus, spleens were collected and
fixed with 10% neutral buffered formalin. The tissues were sectioned and
stained with hematoxylin and eosin. Lesions to the spleen were quantified by
Image-Pro Plus software (v.3.0; Media Cybernetics, Silver Spring, MD).
Specifically, images of lymphocyte follicles were digitized with a Dell
computer (400 MHz) and a Hitachi KP-D50 color camera attached to an
Olympus BH-2 microscope. The Image-Pro counting module was used to
quantify pycnotic lymphocytes, i.e. those lymphocytes exhibiting condensed
and deeply stained chromatin. The best scoring accuracy was obtained by
adjusting the color range (red) upper threshold and the lower thresholds of
the area, density and heterogeneity parameters. For each mouse four fields
(four separate follicles) were analyzed. The average numbers of pycnotic cells
per field were compared between treatment groups by ANOVA (StatView
v.4.5). P ⬍ 0.05 was considered a significant effect.
Dose range finding experiment: in vivo AhR antagonist activity of 3⬘M4⬘NF
To investigate the ability of 3⬘M4⬘NF to antagonize the AhR in vivo, we
utilized male mice heterozygous for a DRE–lacZ transgene (aged 7–12 weeks).
These mice express β-galactosidase (β-gal) in response to AhR agonist
treatment (18). Animals were divided into four treatment groups and were
injected i.p. with vehicle, 15 µg/kg TCDD or 15 µg/kg TCDD plus 0.6 or
6 mg/kg 3⬘M4⬘NF. Note that 3⬘M4⬘NF was administered as a split dose: one
treatment 4 h prior to TCDD, a second concurrently with TCDD and a third
4 h after TCDD (i.e. 0, 0.2 or 2 mg/kg/injection). The vehicle control group
comprised three mice and each of the other three treatment groups comprised
four mice. Sixteen hours after TCDD exposure mice were killed by CO2
asphyxiation and liver and lung sections were collected for β-gal measurements.
Specifically, tissue sections (~10⫻10⫻5 mm) were finely minced and transferred to hypotonic ammonium chloride solution (ACK buffer) to lyse red
blood cells. Minced tissue was subsequently rinsed in phosphate-buffered
saline and transferred to 1 ml of tissue lysis solution containing 100 mM
potassium phosphate, pH 7.8, and 0.2% Triton X-100, supplemented with
phenylmethylsulfonyl fluoride (0.2 mM), leupeptin (1 µg/ml) and dithiothreitol
(1 mM) immediately before use. Tissue was homogenized between the ends
of frosted glass slides and transferred to microcentrifuge tubes. All extracts
were centrifuged at 12 000 g for 5 min and supernatants were held at 48°C
for 60 min to inactivate endogenous β-gal (23). Samples were stored at
–20°C until analysis. Transgene-encoded bacterial β-gal activity was measured
with a commercial luminescence assay system (Galacto Light Plus; Tropix).
Activities were normalized to protein concentration (NanoOrange protein
quantitation kit; Molecular Probes). The effect of treatment on TCDD-induced
β-gal activity was evaluated by ANOVA. P ⬍ 0.05 was considered a
significant effect.
Effect of 3⬘M4⬘NF on B[a]P and CSC genotoxicity
Twenty wild-type C57Bl/6J mice, age 7–8 weeks, were randomly assigned to
five treatment groups: vehicle control and B[a]P plus 0, 0.2, 2 or 20 mg/kg/
injection 3⬘M4⬘NF. B[a]P was delivered as a bolus of 150 mg/kg i.p. and
3⬘M4⬘NF was administered according to our standard split dose schedule, 4 h
prior to, concurrently with and 4 h after the B[a]P challenge. Fifty-four hours
post-B[a]P treatment peripheral blood samples were collected from the tail
vein of each mouse. Mice were killed by CO2 overdose and spleens were
weighed and fixed in 10% neutral buffered formalin. Blood cells were collected
and processed for flow cytometric analysis of genotoxicity as described below.
Spleens were embedded, sectioned and stained for analysis of pycnotic
lymphocytes as described.
In a second experiment, 15 wild-type C57Bl/6J mice (age 7–8 weeks) were
divided among three treatment groups: vehicle control, CSC and CSC plus
3⬘M4⬘NF. CSC was administered at 24 h intervals for 3 consecutive days,
25 mg/kg/injection. 3⬘M4⬘NF was administered according to the split dose
schedule described above, 4 h prior to, concurrently with and 4 h after each
CSC challenge. The concentration of 3⬘M4⬘NF that we chose was the lowest
in vivo concentration that was studied in the DRE–lacZ and B[a]P co-exposure
models, 0.2 mg/kg/injection (i.e. 0.6 mg/kg/day). Twenty-four hours after the
last CSC treatment peripheral blood and spleen tissues were prepared as
described above.
Genotoxicity measurements
The frequency of micronuclei in peripheral blood reticulocytes was measured
to index cytogenetic damage. Micronuclei arise as a consequence of clastogenic
or aneugenic action and this end-point is widely used to evaluate the
carcinogenic potential of test agents (24,25). For the current study a high
throughput flow cytometric technique was used to quantify the incidence of
micronuclei. Heparinized blood from each animal was fixed with ultra-cold
methanol and prepared for analysis as described previously (26). Briefly, cells
were collected by centrifugation and incubated in flow cytometry tubes with
80 µl of working RNase solution plus anti-CD71–FITC (1 mg RNase and 10
µl stock anti-CD71–FITC per ml bicarbonate-buffered saline). Cells were
AhR signaling and cytogenetic damage
Table I. Effect of 3⬘M4⬘NF on zoxazolamine paralysis time
Pre-treatment
Conc.
(mg/kg)
n
Mean zoxazolamine
paralysis timea (⫾SD)
Solvent
3⬘M4⬘NF
3⬘M4⬘NF
3⬘M4⬘NF
3⬘M4⬘NF
0
0.2
2
20
40
14
5
10
10
4
1.00
1.01
1.03
1.44
6.82
⫾
⫾
⫾
⫾
⫾
0.29
0.15
0.20
0.47
4.57
P valueb
0.902
0.702
0.009*
⬍0.0001*
aNormalized
to mean paralysis time observed for the solvent control group
for each experiment (set to 1.0).
t-test results (two tailed), comparing average paralysis time with
that observed for the solvent control group. Asterisks denote statistical
significance, P ⬍ 0.05.
bUnpaired
maintained at 4°C for 30 min, moved to room temperature for 30 min and
then kept on ice until analysis (the same day). Immediately before each
sample was analyzed 1 ml of ice-cold PI solution was added (1.25 µg/ml PI
in bicarbonate-buffered saline).
Flow cytometric analysis was accomplished with a Becton Dickinson
FacStarPlus tuned to provide 488 nm excitation. The FL1 detector was
configured with a bandpass filter to collect emissions between 520 and 560
nm. PI-associated fluorescence was collected in the FL2 channel with a 580
long pass filter. Before scoring samples the instrument was calibrated with
anti-CD71 and PI stained malaria-infected erythrocytes, which guide instrument settings including photomultiplier tube voltages and compensation values
(27). After calibration each sample was scored for reticulocyte (RET) and
micronucleated reticulocyte (MN-RET) frequency upon acquisition of 20 000
total RETs (CellQuest v.3.0.1 software; Becton Dickinson). ANOVA was
performed to evaluate the effect of treatment on mean RET and MN-RET
frequencies. P ⬍ 0.05 was considered a significant effect.
Results
Inhibition of CYP1A activity in vivo
Initial experiments were designed to define in vivo doses of
3⬘M4⬘NF that are relatively specific for AhR antagonist activity
and included a functional assay of CYP1A inhibition, the
zoxazolamine paralysis test. Paralysis times, which are
inversely related to CYP1A activity, are presented in Table I.
Pre-treatment of mice with 3⬘M4⬘NF at 20 and 40 mg/kg
significantly increased mean paralysis time. Conversely, no
measurable effect was observed at 0.2 and 2 mg/kg 3⬘M4⬘NF
relative to vehicle controls. These results are consistent with
previous observations that this flavonoid inhibits EROD activity in vitro (12,15) and suggest that low concentrations of
3⬘M4⬘NF have no significant effect on the basal level of
CYP1A1/2 enzyme activity in vivo.
Ahr null allele mice and splenic lesions
In a further effort to assess 3⬘M4⬘NF for its impact on nonAhR targets, male Ahr null allele mice were treated with
vehicle, 150 mg/kg B[a]P or 150 mg/kg B[a]P plus 0.2, 2 or
20 mg/kg/injection 3⬘M4⬘NF (three injections). We utilized
Ahr null allele mice in this experiment because they are more
sensitive to the combined toxicity of B[a]P and 3⬘M4⬘NF (8).
Furthermore, the toxicity evident in these animals would by
definition be related to non-AhR targets. After 48 h spleens
were collected, fixed, sectioned and stained with hematoxylin
and eosin. Consistent with previous experiments (15), mice
co-treated with B[a]P and 20 mg/kg/injection 3⬘M4⬘NF (three
injections) exhibited small spleens with reduced red pulp.
Additionally, the incidence of pycnotic spleen lymphocytes
was highly elevated (Figure 1). The number of pycnotic
lymphocytes per high power field were scored with ImagePro Plus software and the results are presented in Figure 2. A
Fig. 1. Section of spleen from a male Ahr–/– mouse treated with 150 mg/kg
B[a]P (A) or co-treated with 150 mg/kg B[a]P plus 20 mg/kg/injection
3⬘M4⬘NF, with three injections (B). Note the high percentage of
lymphocytes which exhibit condensed, pycnotic chromatin in B. This is a
lesion which is not observed in mice treated with 3⬘M4⬘NF alone or even
B[a]P alone at this relatively well-tolerated dose.
marked increase in spleen lesions was found in the B[a]P plus
high 3⬘M4⬘NF treatment group. On the other hand, this
toxicity was not evident at lower concentrations of flavonoid.
Importantly, these results correlate with data obtained with
zoxazolamine paralysis times. While 20 mg/kg 3⬘M4⬘NF
inhibits CYP1A activity, the lower concentrations of 0.2 and
2 mg/kg do not result in a measurable effect. These data
suggest that the splenic toxicity of B[a]P plus 3⬘M4⬘NF is a
result of inhibition of constitutive CYP1A activity resulting in
either a change in deposition, clearance and/or bioactivation
of the procarcinogen.
In vivo AhR antagonist activity of 3⬘M4⬘NF
Data from the zoxazolamine and the Ahr null allele mouse
experiments suggest that injections of 3⬘M4⬘NF in the range
0.2–2 mg/kg may minimally affect non-AhR target(s), and
CYP1A activity in particular. To test the ability of 3⬘M4⬘NF
to block AhR-mediated signaling in vivo, these concentrations
were evaluated in a transgenic mouse model that responds to
AhR agonists by the induction of lacZ. Heterozygous mice
173
S.D.Dertinger et al.
Fig. 2. The mean number of pycnotic lymphocytes per high magnification
field are graphed for each of five treatment groups. Standard deviation bars
are included. B[a]P, 150 mg/kg; 0.2 mg 3⬘M4⬘NF, 0.2 mg/kg/injection, three
injections; 2 mg 3⬘M4⬘NF, 2 mg/kg/injection, three injections; 20 mg
3⬘M4⬘NF, 20 mg/kg/injection, three injections. Scoring was of four
randomly chosen follicles per mouse. The Image-Pro count module was
used to quantify the number of pycnotic cells per field. The effect of
treatment on the mean number of pycnotic cells per field was evaluated by
ANOVA and Fisher’s PLSD test. The asterisk denotes a statistically
significant difference in mean number of pycnotic cells per field compared
with each of the other four treatment groups, P ⬍ 0.05.
were treated with 15 µg/kg TCDD and 0, 0.2 or 2 mg/kg
3⬘M4⬘NF (three injections). After 16 h liver and lung tissues
were processed for β-gal measurements. These enzyme activities, normalized to protein concentration, are presented in
Figure 3.
As expected, mean β-gal activity was markedly enhanced
in the TCDD treatment group. Administration of the split
dose of 3⬘M4⬘NF reduced this activity. Although statistical
significance was not achieved in the liver, a clear dosedependent reduction was observed. In fact, at a total dose of
6 mg/kg 3⬘M4⬘NF we saw a nearly complete inhibition of
TCDD-induced activity. As observed in other experiments
(data not shown), the liver was somewhat variable in terms of
TCDD-induced β-gal activity, and this variation is likely
responsible for the lack of statistical significance in this
compartment. One possible source of variation is the relatively
high hemoglobin content associated with the liver extracts (we
have observed that hemoglobin interferes with chemiluminescence-based β-gal assays; (28). TCDD-induced β-gal activity
was less variable in the lung. Although 3⬘M4⬘NF was not
shown to reduce transgene expression to as great an extent as
in the liver, the dose-dependent reduction was statistically
significant (ANOVA, P ⫽ 0.046). These data directly demonstrate the in vivo AhR antagonist activity of the flavonoid
3⬘M4⬘NF in the concentration range 0.2–2 mg/kg/injection.
Effect of 3⬘M4⬘NF on B[a]P and CSC genotoxicity
With a better understanding of the 3⬘M4⬘NF concentrations
which are more specific for AhR antagonism compared with
the one high dose that we initially studied, we were in a better
position to evaluate the influence of AhR-dependent signal
174
Fig. 3. Mean β-gal activity of liver and lung tissue (from DRE–lacZ
transgenic mice) is graphed for each of four treatment groups. Standard
deviation bars are included. β-Gal activity induced by 15 µg/kg TCDD is
antagonized in a dose-dependent fashion by 3⬘M4⬘NF (0.2 mg 3⬘M4⬘NF,
0.2 mg/kg/injection, three injections; 2 mg 3⬘M4⬘NF, 2 mg/kg/injection,
three injections). ANOVA results comparing β-gal activity among groups
treated with TCDD: liver, P ⫽ 0.140; lung, P ⫽ 0.046.
transduction on the genotoxicity of cigarette smoke mutagens.
In one experiment five male C57Bl/6J mice per group were
treated with vehicle, B[a]P alone or B[a]P plus 0.6 or 6 mg/
kg 3⬘M4⬘NF (0.2 or 2 mg/kg, three injections).
Observations over the 54 h treatment period suggest that at
the concentrations tested 3⬘M4⬘NF did not potentiate B[a]P
toxicity. The animals did not exhibit any obvious signs of
distress, i.e. the mice were able to maintain caloric intake,
retain their righting reflex, etc. Furthermore, spleen weight
was not affected by B[a]P and 3⬘M4⬘NF nor was the frequency
of pycnotic lymphocytes enhanced (data not shown). These
results are consistent with data from the experiment with the
highly sensitive Ahr null allele mice in which 20 but not 2 or
0.2 mg/kg/injection 3⬘M4⬘NF potentiated certain B[a]P lesions.
Collectively, these data suggest that at the concentrations tested
3⬘M4⬘NF does not severely affect that non-AhR target(s)
which is responsible for the synergistic toxicity that can
result when high concentrations of 3⬘M4⬘NF are combined
with B[a]P.
AhR signaling and cytogenetic damage
Table II. Effect of 3⬘M4⬘NF on CSC toxicity
Fig. 4. Attenuation of B[a]P-induced toxicity by 3⬘M4⬘NF. Mean RET and
MN-RET frequencies are shown. Five animals per group; error bars indicate
1 SD. B[a]P, 150 mg/kg; 0.2 mg 3⬘M4⬘NF, 0.2 mg/kg/injection, three
injections; 2 mg 3⬘M4⬘NF, 2 mg/kg/injection, three injections. Two tailed ttest results: *P ⬍ 0.0001, MN-RET (%) significantly different to solvent
control; **P ⬍ 0.01, RET (%) significantly different to solvent control; †P
⬍ 0.01, MN-RET (%) significantly different to B[a]P only group; ††P ⬍
0.01, RET (%) significantly different to B[a]P only group; #P ⬍ 0.001,
MN-RET (%) significantly different to B[a]P only group; ##P ⬍ 0.005, RET
(%) significantly different to B[a]P only group.
The effect of 3⬘M4⬘NF treatment on B[a]P genetic toxicity
is presented in Figure 4. As these data clearly indicate, the
flavonoid is capable of attenuating B[a]P-induced chromosome
damage. Furthermore, B[a]P-induced stem cell toxicity, as
measured by reticulocyte frequency, was abolished by
3⬘M4⬘NF co-treatment. Both of these protecting effects
were dose dependent and found to be statistically significant
(P ⬍ 0.05).
This experiment was followed by a study in which C57Bl/
6J mice were treated with CSC with and without 3⬘M4⬘NF
co-treatment. By substituting the complex mixture of chemicals
in CSC for B[a]P we hoped to more realistically model
exposure to tobacco smoke. We have previously found that a
bolus of 25 mg/kg CSC does not induce a measurable level
of DNA damage as measured by micronucleus formation,
while a repeat dosing schedule does (data not shown). Therefore, these experiments necessarily involved repeat dosing:
CSC administered once a day for 3 days. As with previous
experiments, we bracketed each CSC dose with 3⬘M4⬘NF
injections.
The animals were closely monitored for acute signs of
toxicity. No outward signs of distress were evident. Furthermore, spleen weights and the frequency of pycnotic lymphocytes in the spleen compartment were not affected by the CSC
or CSC plus 3⬘M4⬘NF treatments. Stem cell toxicity and
genotoxicity data are provided in Table II. While there have
been reports that high concentrations of CSC can arrest mitosis
of cultured cells (29), we did not observe any change in
erythropoiesis function among treatment groups, i.e. no significant difference was observed in RET frequencies. On the other
hand, the incidence of MN-RET was significantly elevated in
the CSC treatment group over the vehicle control. This low
level of clastogenic activity associated with CSC was abrogated
by the AhR antagonist 3⬘M4⬘NF.
Mouse no.
Sex
Treatment
RETa (%)
MN-RETb (%)
1
2
3
4
5
Average
SD
6
7
8
9
10
Average
SD
11
12
13
14
15
Average
SD
M
M
M
M
M
Vehicle
Vehicle
Vehicle
Vehicle
Vehicle
M
M
M
M
M
CSCc
CSC
CSC
CSC
CSC
M
M
M
M
M
CSC
CSC
CSC
CSC
CSC
2.34
1.89
2.21
1.84
1.94
2.04
0.22
2.87
1.98
1.67
1.80
2.53
2.17
0.51
2.25
1.96
2.53
1.87
1.64
2.05
0.35
0.38
0.37
0.38
0.37
0.35
0.37
0.01
0.70
0.51
0.48
0.61
0.44
0.55d
0.11
0.35
0.34
0.42
0.38
0.37
0.37
0.03
⫹
⫹
⫹
⫹
⫹
3⬘M4⬘NFe
3⬘M4⬘NF
3⬘M4⬘NF
3⬘M4⬘NF
3⬘M4⬘NF
aRET (%),
bMN-RET
reticulocyte frequency, a measure of bone marrow cytotoxicity.
(%), micronucleated reticulocyte frequency, a measure of
genotoxicity.
cCSC, cigarette smoke condensate, administered at 24 h intervals for 3 days,
25 mg/kg/day.
dEffect of treatment on mean MN-RET frequency was evaluated by
ANOVA, P ⫽ 0.0011; Fisher’s PLSD test indicates that the mean MN-RET
frequency of the CSC group is significantly different to the solvent control
and the CSC ⫹ 3⬘M4⬘NF groups.
eCSC ⫹ 3⬘M4⬘NF, CSC administered at 24 h intervals for 3 days,
25 mg/kg/day; 3⬘M4⬘NF administered 4 h prior to, concurrently with and
4 h after CSC at 0.2 mg/kg/injection.
Discussion
Cigarette smoke contains compounds which can transform the
cytosolic AhR to an active transcription factor. However,
the consequences of AhR-mediated enzyme induction on the
toxicity of cigarette smoke are not clear. For instance, with
regard to DNA damaging capacity it is possible that constitutive
cytochrome P450 activities dictate the potency of the mixture
or that other biochemical processes (e.g. DNA repair) obscure
any differences resulting from AhR-mediated gene transcription. Alternately, AhR-dependent events may significantly
influence cigarette smoke-induced DNA damage. This information is potentially important, for it may provide the framework
for identifying highly sensitive individuals and may represent
an approach for attenuating the toxicity of similar types of
exposures.
One of the extreme difficulties associated with the study of
host and environmental factors which influence cigarette
smoke-induced toxicity and carcinogenicity is that exposure
to tobacco smoke is so complex. Approximately 4000 chemicals are found in tobacco smoke and over 50 of these
compounds are recognized as known or probable carcinogens
(30). Given the multitude of bioactive compounds in smoke,
it is clear that investigations hoping to supply new, mechanistic
information must employ very refined and specific methods.
In studying the influence of AhR-mediated events the availability of AhR mutant cells lines, DRE–reporter cells and animals
and Ahr null allele mice are important and necessary tools.
Even so, it is clear that these tools alone cannot always
discriminate between the effects that agonist-induced up175
S.D.Dertinger et al.
regulation of DRE-controlled genes have on end-points of
toxicity from the consequences that a complete lack of AhR
has on setting certain basal enzyme levels such as CYP1A2
(11). For the experiments described herein we successfully
employed 3⬘M4⬘NF to isolate and study the influence that
enzyme induction through AhR signaling has on the genotoxicity of B[a]P and CSC in vivo. This information is potentially
important, because it is presumably more relevant to actual
exposure conditions compared with models in which signaling
has been permanently ablated and constitutive metabolism
profiles have therefore been markedly altered.
Collectively, the data presented herein support results from
earlier experiments with AhR mutant cell lines and Ahr null
allele mice which suggest that AhR signaling may have a net
potentiating effect on the toxicity, especially the genotoxicity,
of cigarette smoke constituents. Experimentally, we have
demonstrated that the flavone derivative 3⬘M4⬘NF can be used
at low concentrations which do not significantly affect the
constitutive in vivo activity of P450 CYP1A but which retain an
ability to antagonize AhR signaling. In this same concentration
range 3⬘M4⬘NF was found to protect mice from B[a]P- and
CSC-induced cytogenetic damage. These results therefore
strongly suggest that AhR signaling plays an important role
in mediating the genetic toxicity of cigarette smoke. A more
complete understanding of the significance of AhR signaling
on cigarette smoke genotoxicity, and ultimately tumorigenicity,
will come from additional studies which incorporate other
end-points of DNA damage. Furthermore, if these additional
genetic toxicity measurements are obtained in different tissue
compartments, for example the lung, added relevance may be
achieved.
Although we report here that 3⬘M4⬘NF is an effective AhR
antagonist in vivo and efficacy can be achieved with some
degree of specificity, it is important to realize that other
potential biological activities of the synthetic flavonoid have
not been fully characterized. For instance, bioflavonoids and
various flavone derivatives have been found to exhibit antioxidant activity (31), induce apoptosis (32), suppress cell cycle
progression (33) and inhibit mitogen-activated protein kinase
kinase (34). Therefore, experiments designed to temporally
and specifically block AhR signal transduction may benefit
from further characterization of 3⬘M4⬘NF and other flavones
for other (non-AhR) biological activities which may be partially
responsible for the protective effects reported here and elsewhere (35–40).
Acknowledgements
The authors wish to thank the technicians, students and post-doctoral staff in
the Gasiewicz laboratory for their critical review of this manuscript. The
authors especially want to thank Andrew Kende for synthesizing 3⬘M4⬘NF, J.
Jeffrey Willey for generating and initially characterizing DRE–lacZ transgenic
animals, Frank Gonzalez and Pedro Fernandez-Salguero for supplying the
founder Ahr null allele mice, Denise Hahn, Nancy Fiore and Cheryl Hurley
for the expert animal care they provided and Litron Laboratories for use of
their flow cytometer. This work was supported by NIH Grants ES02515,
ES09430, ES04862, ES09702 and ES07216, Center Grant ES01247 and
Training Grant ES07026.
References
1. Löfroth,G. and Rannug,A. (1988) Ah receptor ligands in tobacco smoke.
Toxicol. Lett., 42, 131–136.
2. Gebremichael,A., Tullis,K., Denison,M.S., Cheek,J.M. and Pinkerton,K.E.
(1996) Ah-receptor-dependent modulation of gene expression by aged and
diluted sidestream cigarette smoke. Toxicol. Appl. Pharmacol., 141, 76–83.
176
3. Graziano,M.J. and Dorogh,H.W. (1984) Effect of cigarette smoking on
hepatic biotransformations in rats. Toxicol. Appl. Pharmacol., 75, 229–239.
4. Guengerich,F.P. (1991) Reactions and significance of cytochrome P450
enzymes. J. Biol. Chem., 266, 10019–10022.
5. Rowlands,J. and Gustafsson,J. (1997) Aryl hydrocarbon receptor-mediated
signal transduction. Crit. Rev. Toxicol., 27, 109–134.
6. Kellermann,G., Shaw,C.R. and Luyten-Kellerman,M. (1973) Aryl
hydrocarbon hydroxylase inducibility and bronchogenic carcinoma.
N. Engl. J. Med., 289, 934–937.
7. Kouri,R.E., McKinney,E., Slomiany,D.J., Snodgrass,D.R., Wray,N.P. and
McLemore,T.L. (1982) Positive correlation between high aryl hydrocarbon
hydroxylase activity and primary lung cancer as analyzed in cryopreserved
lymphocytes. Cancer Res., 42, 5030–5037.
8. Dertinger,S.D., Silverstone,A.E. and Gasiewicz,T.A. (1998) Influence of
aromatic hydrocarbon receptor-mediated events on the genotoxicity of
cigarette smoke condensate, Carcinogenesis, 19, 2037–2042.
9. Shimizu,Y., Nakatsuru,Y., Ichinose, M,. Takahashi,Y., Kume,H., Mimura,J.,
Fujii-Kuriyama,Y. and Ishikawa,T. (2000) Benzo[a]pyrene carcinogenicity
is lost in mice lacking the aryl hydrocarbon receptor. Proc. Natl Acad.
Sci. USA, 97, 779–782.
10. Ott,M.G. and Zober,A. (1996) Morbidity study of extruder personnel with
potential exposure to brominated dioxins and furans. 2. Results of clinical
laboratory studies. Occup. Environ. Med., 53, 844–846.
11. Lahvis,G.P. and Bradfield,C.A. (1998) Ahr null alleles: distinctive or
different? Biochem. Pharmacol., 56, 781–787.
12. Lu,Y.F., Santostefano,M., Cunningham,B.D.M., Threadgill,M.D. and
Safe,S. (1995) Identification of 3⬘-methoxy-4⬘-nitroflavone as a pure aryl
hydrocarbon (Ah) receptor antagonist and evidence for more than one
form of the nuclear Ah receptor in MCF-7 human breast cancer cells.
Arch. Biochem. Biophys., 316, 470–477.
13. Gasiewicz,T.A., Kende,A.S., Rucci,G., Whitney,B. and Willey,J.J. (1996)
Analysis of structural requirements for Ah receptor antagonist activity:
ellipticines, flavones and related compounds. Biochem. Pharmacol., 52,
1787–1803.
14. Henry,E.C., Kende,A.S., Rucci,G., Totleben,J., Willey,J.J., Dertinger,S.D.,
Pollenz,R.S., Jones,J.P. and Gasiewicz,T.A. (1999) Flavone antagonists
bind competitively with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) to
the aryl hydrocarbon receptor but inhibit nuclear uptake and transformation.
Mol. Pharmacol., 55, 716–725.
15. Dertinger,S.D., Lantum,H.B.M., Silverstone,A.E. and Gasiewicz,T.A.
(2000) Effect of 3⬘-methoxy-4⬘-nitroflavone on benzo[a]pyrene toxicity:
aryl hydrocarbon receptor-dependent and -independent mechanisms.
Biochem. Pharmacol., 60, 189–196.
16. Cunningham,B.D.M., Threadgill,M.D., Groundwater,P.W., Dale,I.L. and
Hickman,J.A. (1992) Synthesis and biological evaluation of a series of
flavones designed as inhibitors of protein tyrosine kinases. Anticancer
Drug Des., 7, 365–384.
17. Fernandez-Salguero,P., Pineau,T., Hilbert,D.M., McPhail,T., Lee,S.S.T.,
Kimma,S., Nebert,D.W., Rudikoff,S., Ward,J.M. and Gonzalez,F.J. (1995)
Immune system impairment and hepatic fibrosis in mice lacking the
dioxin-binding Ah receptor. Science, 268, 722–726.
18. Willey,J.J., Stripp,B.R., Baggs,R.B. and Gasiewicz,T.A. (1998) Ah receptor
activation in genital tubercle, palate and other embryonic tissues in TCDDresponsive lacZ mice. Toxicol. Appl. Pharmacol., 151, 33–44.
19. Conney,A.H., Davison,C., Gastel,R. and Burns,J.J. (1960) Adaptive
increase in drug metabolizing enzymes induced by phenobarbital and other
drugs. J. Pharmacol. Exp. Ther., 130, 1–8.
20. Robinson,J.R. and Nebert,D.W. (1974) Genetic expression of aryl
hydrocarbon hydroxylase induction. Presence or absence of association
with zoxazolamine, diphenylhydantoin and hexobarbital metabolism. Mol.
Pharmacol., 10, 484–493.
21. Atal,C.K., Dubey,R.K. and Singh,J. (1985) Biochemical basis of enhanced
drug bioavailability by piperine: evidence that piperine is a potent inhibitor
of drug metabolism. J. Pharmacol. Exp. Ther., 232, 258–262.
22. Vancutsem,P.M. and Babish,J.G. (1993) Effects of ciprofloxacin and
enrofloxacin on zoxazolamine kinetics, plasma concentration and sleeping
times in mice. Toxicol. Lett., 69, 1–14.
23. Young,D.C., Kingsley,S.D., Ryan,K.A. and Dutko,F.J. (1993) Selective
inactivation of eukaryotic beta-galactosidase in assays for inhibitors of
HIV-1 TAT using bacterial beta-galactosidase as a reporter enzyme. Anal.
Biochem., 215, 24–30.
24. Hayashi,M., Tice,R.R., MacGregor,J.T., Anderson,D., Blakey,D.H., KirschVolders,M., Oleson,F.B.Jr, Pacchierotti,F., Romagna,F., Shimada,H.,
Sutou,S. and Vannier,B. (1994) In vivo rodent erythrocyte micronucleus
assay. Mutat. Res., 312, 293–304.
25. Hayashi,M., MacGregor,J.T., Gatehouse,D.G., Adler,I.-D., Blakey,D.H.,
Dertinger,S.D., Krishna,G., Morita,T., Russo,A. and Sutou,S. (2000) In
AhR signaling and cytogenetic damage
vivo rodent erythrocyte micronucleus assay. II. Some aspects of protocol
design including repeat treatments, integration with toxicity testing and
automated scoring. Environ. Mol. Mutagen., 35, 234–252.
26. Dertinger,S.D., Torous,D.K. and Tometsko,K.R. (1996) Simple and reliable
enumeration of micronucleated reticulocytes with a single-laser flow
cytometer. Mutat. Res., 371, 283–292.
27. Dertinger,S.D., Torous,D.K., Hall,N.E., Tometsko,C.R. and Gasiewicz,T.A.
(2000) Malaria-infected erythrocytes serve as biological standards to
ensure reliable and consistent scoring of micronucleated erythrocytes by
flow cytometry. Mutat. Res., 464, 195–200.
28. Nazarenko,D.A., Dertinger,S.D., Gasiewicz,T.A. Enhanced detection of β.
galactosidase reporter activation by a luminescent assay in tissue lysates
is achieved by a reduction of hemoglobin content. Biotechniques. In press.
29. Hsu,T.C., Cherry,L.M., Bucana,C., Shirley,L.R. and Gairola,C.G. (1991)
Mitosis-arresting effects of cigarette smoke condensate on human lymphoid
cell lines. Mutat. Res., 259, 67–78.
30. National Cancer Institute (1999) Smoking and Tobacco Control Monograph
No. 10. US Department of Health and Human Services, Public Health
Service, National Institutes of Health, Bethesda, MD.
31. Kono,Y., Kobayashi,K., Tagawa,S., Adachi,K., Ueda,A., Sawa,Y. and
Shibata,H. (1997) Antioxidant activity of polyphenolics in diets. Rate
constants of reactions of chlorogenic acid and caffeic acid with reactive
species of oxygen and nitrogen. Biochim. Biophys. Acta, 1335, 335–342.
32. Csokay,B., Prajda,N., Weber,G. and Olah,E. (1997) Molecular mechanisms
in the antiproliferative action of quercetin. Life Sci., 60, 2157–2163.
33. Reiners,J.Jr, Clift,R. and Mathieu,P. (1999) Suppression of cell cycle
progression by flavonoids: dependence on the aryl hydrocarbon receptor.
Carcinogenesis, 20, 1561–1566.
34. Reiners,J.Jr, Lee,J.Y., Clift,R., Dudley,D. and Myrand,S. (1998) PD98059
is an equipotent antagonist of the aryl hydrocarbon receptor and inhibitor
of mitogen-activated protein kinase kinase. Mol. Pharmacol., 53, 438–445.
35. Lamartiniere,C., Moore,J., Brown,N., Thompson,R., Hardin,M. and
Barnes,S. (1995) Genistein suppresses mammary cancers in rats.
Carcinogenesis, 16, 2833–2840.
36. Verma,A., Johnson,J., Gould,M. and Tanner,M. (1988) Inhibition of 7,12dimethylbenz[a]anthraceneand N-nitrosomethylurea-induced
rat
mammary cancer by dietary flavonol quercetin. Cancer Res., 48, 5754–
5758.
37. Heo,M.Y., Yu,K.S., Kim,K.H., Kim,H.P. and Au,W.W. (1992) Anticlastogenic effect of flavonoids against mutagen-induced micronuclei in
mice. Mutat. Res., 284, 243–249.
38. Choi,J.S., Park,K.Y., Moon,S.H., Rhee,S.H. and Young,H.S. (1994)
Antimutagenic effect of plant flavonoids in the Salmonella assay system.
Arch. Pharm. Res., 17, 71–75.
39. Guengerich,F.P. and Kim,D.H. (1990) In vitro inhibition of dihydropyridine
oxidation and aflatoxin B1 activation in human liver microsomes by
niringenin and other flavonoids. Carcinogenesis, 11, 2275–2279.
40. Elangovan,V., Sekar,N. and Govindasamy,S. (1994) Chemopreventive
potential of dietary bioflavonoids against 20-methylcholanthrene-induced
tumorigenesis. Cancer Lett., 87, 107–113.
Received August 15, 2000; revised October 17, 2000,
accepted October 20, 2000
177