Toxicology 325 (2014) 12–20

Contents lists available at ScienceDirect

Toxicology
journal homepage: www.elsevier.com/locate/toxicol

Genotoxic, epigenetic, and transcriptomic effects of tamoxifen in

mouse liver$
Aline de Conti a,1, Volodymyr Tryndyak a,1, Mona I. Churchwell a , Stepan Melnyk b ,
John R. Latendresse c , Levan Muskhelishvili c , Frederick A. Beland a , Igor P. Pogribny a, *
a
Division of Biochemical Toxicology, National Center for Toxicological Research, Jefferson, AR 72079, USA
b
Department of Pediatrics, University of Arkansas for Medical Sciences, Little Rock, AR 72202, USA
c
Toxicologic Pathology Associates, National Center for Toxicological Research, FDA, Jefferson, AR 72079, USA

A R T I C L E I N F O A B S T R A C T

Article history: Tamoxifen is a non-steroidal anti-estrogenic drug widely used for the treatment and prevention of breast
Received 30 May 2014 cancer in women; however, there is evidence that tamoxifen is hepatocarcinogenic in rats, but not in
Received in revised form 16 July 2014 mice. Additionally, it has been reported that tamoxifen may cause non-alcoholic fatty liver disease
Accepted 10 August 2014
(NAFLD) in humans and experimental animals. The goals of the present study were to (i) investigate the
Available online 12 August 2014
mechanisms of the resistance of mice to tamoxifen-induced hepatocarcinogenesis, and (ii) clarify effects
of tamoxifen on NAFLD-associated liver injury. Feeding female WSB/EiJ mice a 420 p.p.m. tamoxifen-
Keywords:
containing diet for 12 weeks resulted in an accumulation of tamoxifen-DNA adducts, (E)-
Tamoxifen
Mouse
a-(deoxyguanosin-N2-yl)-tamoxifen (dG-TAM) and (E)-a-(deoxyguanosin-N2-yl)-N-desmethyltamox-
Liver ifen (dG-DesMeTAM), in the livers. The levels of hepatic dG-TAM and dG-DesMeTAM DNA adducts in
DNA adducts tamoxifen-treated mice were 578 and 340 adducts/108 nucleotides, respectively, while the extent of
Epigenetics global DNA and repetitive elements methylation and histone modifications did not differ from the values
Gene expression in control mice. Additionally, there was no biochemical or histopathological evidence of NAFLD-
associated liver injury in mice treated with tamoxifen. A transcriptomic analysis of differentially
expressed genes demonstrated that tamoxifen caused predominantly down-regulation of hepatic lipid
metabolism genes accompanied by a distinct over-expression of the lipocalin 13 (Lcn13) and peroxisome
proliferator receptor gamma (Pparg), which may prevent the development of NAFLD. The results of
the present study demonstrate that the resistance of mice to tamoxifen-induced liver carcinogenesis may
be associated with its ability to induce genotoxic alterations only without affecting the cellular
epigenome and an inability of tamoxifen to induce the development of NAFLD.
Published by Elsevier Ireland Ltd.

1. Introduction

Abbreviations: NAFLD, Non-alcoholic fatty liver disease; dG-TAM, (E)-a-(deox-
yguanosin-N2-yl)tamoxifen; dG-DesMeTAM, (E)-a-(deoxyguanosin-N2-yl)-N-des-
methyltamoxifen; ALT, alanine transaminase; AST, aspartate transaminase; LDH,
lactate dehydrogenase; a-SMA, a-smooth muscle actin; Col1a1, collagen, type I,
alpha 1; Lcn13, lipocalin 13; Pparg, peroxisome proliferator receptor gamma;
Dnmt1, DNA methyltransferase 1; Dnmt3a, DNA methyltransferase 3a; Dnmt3b,
DNA methyltransferase 3b; SAM, S-adenosyl-L-methionine; SAH, S-adenosyl-L-
homocysteine; LINE 1, long interspersed elements 1; SINE B1, short interspersed
nuclear elements B1; SINE B2, short interspersed nuclear elements B2; H3K4,
histone H3 lysine 4; H3K9, histone H3 lysine 9; H3K27, histone H3 lysine 27; H3K79,
histone H3 lysine 79; H4K20, histone H4 lysine 20; Gapdh, glyceraldehyde-3-
phosphate dehydrogenase; qRT-PCR, quantitative reverse transcription-PCR.
$
The views expressed in this paper do not necessarily represent those of the U.S.
Food and Drug Administration.
* Corresponding author. Tel.: +1 870 543 7096.
E-mail address: igor.pogribny@fda.hhs.gov (I.P. Pogribny).
1
Both authors contributed equally to this work.
Tamoxifen, a non-steroidal anti-estrogenic drug, is commonly
used to prevent the re-occurrence of breast cancer or its
occurrence in healthy women at high risk of developing breast
cancer. Despite the indisputable benefits of tamoxifen, a number of
clinical reports have demonstrated that the use of tamoxifen
increases the incidence of endometrial cancer (Fisher et al., 1994;
Jones et al., 2012) and induces NAFLD (Bruno et al., 2005; Murata
et al., 2000; Oien et al., 1999; Saphner et al., 2009) in humans.
Additionally, it has been demonstrated that tamoxifen is a potent
hepatocarcinogen in rats with both initiating and promoting
activities. This was evidenced by results of several independent
studies showing that long-term tamoxifen exposure induces a
formation of various types of neoplastic lesions in the livers of
different rat strains (Davies et al., 1997; Greaves et al., 1993).

http://dx.doi.org/10.1016/j.tox.2014.08.004
0300-483X/ Published by Elsevier Ireland Ltd.
 A. de Conti et al. / Toxicology 325 (2014) 12–20 13

It is widely believed that the mechanism of tamoxifen-induced
rat hepatocarcinogenesis is associated with various genotoxic
alterations. Tamoxifen and its N-demethylated derivative N-
desmethyltamoxifen are metabolized in the liver through
a-hydroxylation by cytochrome P450-dependent monooxyge-
nases (Kiyotani et al., 2012). The a-hydroxylated metabolites are
subsequently esterified (e.g., sulfation) to reactive derivatives
capable of interacting with DNA to cause the formation of (E)-
a-(deoxyguanosin-N2
yl)-tamoxifen (dG-TAM) and (E)-a-(deox-
yguanosin-N2-yl)-N-desmethyltamoxifen (dG-DesMeTAM) DNA
adducts. Previously, we demonstrated that hepatocarcinogenic
activity of tamoxifen in female rats is associated with its ability to
induce both genotoxic and epigenetic abnormalities in the liver
(Tryndyak et al., 2006, 2007).
In contrast to a well-established hepatocarcinogenicity of
tamoxifen in rats, two reports by Tucker et al. (1984) and Martin
et al. (1997) have demonstrated that tamoxifen is not hepato-
carcinogenic in mice. However, because of some limitations of
these studies, such as shorter than required 2-year treatment
(Tucker et al., 1984) and small size of experimental group (Martin
et al., 1997), the tamoxifen liver carcinogenicity in mice remains
unclear. Umemoto et al. (2001) has attributed this species-
dependent difference in tamoxifen hepatocarcinogenicity to a
substantially lower level of tamoxifen-DNA adducts in the livers of
mice than in rats. However, there is insufficient knowledge to
clarify these existing hepatocarcinogenicity and hepatotoxicity
discrepancies in mice.
In light of this, the goals of the present study were to (i)
investigate the mechanisms of the resistance of mice to tamoxifen-
induced liver carcinogenesis, and (ii) clarify effects of tamoxifen on
NAFLD-associated liver injury in female mice. The results of our
study demonstrate that treatment of mice with tamoxifen for
12 weeks resulted in a substantial accumulation of tamoxifen-DNA
adducts in the livers. In contrast, tamoxifen treatment did not
affect the status of hepatic epigenome.
These findings suggest that distinct rat and mouse differences
in tamoxifen liver toxicity and carcinogenicity may be attributed to
the absence of tamoxifen-induced epigenetic alterations and to
lack of tamoxifen ability to induce the development of NAFLD-
associated liver injury in mice.
2. Material and methods
2.1. Animals and treatments
WSB/EiJ female mice were obtained from the Jackson Laboratory (Bar Harbor,
ME), housed four per cage in a temperature-controlled (24  C) room with a 12-h
light/dark cycle, and given ad libitum access to water and NIH-31 laboratory diet.
Mice were allocated randomly to receive either NIH-31 diet containing 420 p.p.m.
tamoxifen (Dyets, Bethlehem, PA) or control NIH-31 diet. Diets were stored at 4  C
and given ad libitum, with biweekly replacement. Body weights of the mice were

Fig. 1. Pathology endpoints and biochemical and morphological changes in control mice and mice fed tamoxifen. (A) Body and (B) relative liver weight. (C) Plasma levels of
triglycerides, cholesterol, glucose, and activity of ALT, AST and LDH. (D) Representative images of hematoxylin and eosin (a,b) and Ki-67 staining (c,d) in liver sections. Original
magnification, 20. (E) Liver content of triglycerides and phosphatidylcholine (ng/mg protein). (F) Expression of a-Sma and Col1a1 in the livers as measured by qRT-PCR.
Values are means  S.D. (n = 5). * – Denotes a significant (p < 0.05) difference from the control mice; y – denotes significant (p < 0.001; r = 0.758) trend.
14 A. de Conti et al. / Toxicology 325 (2014) 12–20

Fig. 2. HPLC-ES-MS/MS analyses of tamoxifen DNA adducts. MRM chromatograms of dG-TAM, its d6 labeled internal standard (IS) and dG-DesMethylTAM in (A) standard
containing 25 pg each of dG-TAM and d6-dG-TAM and 11.6 pg of dG-DesMeTAM; (B) control mouse liver DNA (114 mg) with 25 pg IS; (C) tamoxifen treated mouse liver DNA
(14.1 mg) with 25 pg IS and containing 586 adducts/108 nucleotides of dG-TAM and 320 adducts/108 nucleotides of dG-DesMethylTAM.

recorded weekly. Five experimental and five control mice were euthanized by
exsanguination following deep isoflurane anesthesia 12 weeks after diet
initiation.
2.2. Sample collection, biochemical measurements and histological evaluation
Blood was collected at sacrifice into BD Vacutainer EDTA-containing blood
collection tubes (BD Biosciences, Franklin Lakes, NJ) by direct puncture of the heart.
Plasma was isolated by centrifugation at 3000 rpm for 10 min at 4  C and stored at
80  C. The livers were excised, and a slice of the median lobe was fixed in 10% neutral
buffered formalin for 48 h, trimmed, processed, embedded in infiltrating media
(Surgipath1 Formula “R”, Leica Biosystems, Richmond, IL), sectioned at approxi-
mately 5 m, mounted on a glass slide, and stained with hematoxylin and eosin for
histopathological examination. The remaining liver was snap-frozen immediately
in liquid nitrogen and stored at
80  C for subsequent analyses. After 48 h, a slice of
the median lobe of the liver that had been fixed in 10% neutral buffered formalin was
trimmed, processed, and embedded. All experimental procedures were reviewed
and approved by the National Center for Toxicological Research Animal Care and
Use Committee.
N-desmethyltamoxifen (dG-DesMeTAM) by electrospray ionization tandem mass
spectrometry (ES-MS/MS) coupled with online sample preparation and high
performance liquid chromatography (HPLC) as described in Gamboa da Costa et al.
(2003). The limit of detection (LOD) for both the dG-TAM and dG-DesMeTAM was
0.03 adducts/108. The tamoxifen-DNA adduct levels are expressed as adducts/108
nucleotides.
2.5. Determination of hepatic triglyceride and phosphatidylcholine contents
The level of triglycerides was determined by using a Triglyceride Colorimetric
Assay Kit (Cayman Chemical Company, Ann Arbor, MI). The level of phospatidylcho-
line was determined by using a Phosphatidylcholine Assay Kit (Sigma–Aldrich, Saint
Louis, MO).
2.6. Determination of the global DNA methylation status by methylation-sensitive
cytosine extension assay
The extent of global DNA methylation was evaluated using a radiolabeled
deoxycytidine-50 -triphosphate ([3H]dCTP) extension assay (Pogribny et al., 1999).

2.3. Biochemical analyses

Serum triglycerides, total cholesterol, glucose, alanine aminotransferase (ALT), Table 1

aspartate aminotransferase (AST), and lactate dehydrogenase (LDH) were measured Hepatic tamoxifen-DNA adduct levels in mice fed 420 p.p.m. tamoxifen for 12
using an ACE Alera1 Clinical Chemistry System (Alfa Wassermann Inc., West weeks.a
Caldwell, NJ).
Group dG-TAM dG-DesMeTAM Total tamoxifen DNA adducts
2.4. Determination of tamoxifen-DNA adduct levels Control 0.21  0.21 Undetectable 0.21
Tamoxifen 578  200 340  81 918  214
DNA was isolated from mice liver tissue using standard digestion with
a
proteinase K, followed by phenol–chloroform extraction and ethanol precipitation. The tamoxifen-DNA adduct levels are expressed as adducts/108 nucleotides
The DNA samples were hydrolyzed to nucleosides and analyzed for (E)- (mean  S.D., n = 5). The LOD for both the dG-Tam and dG-DesMeTAM was
a-(deoxyguanosin-N2-yl)-tamoxifen (dG-TAM) and (E)-a-(deoxyguanosin-N2-yl)- 0.03 adducts/108.
 A. de Conti et al. / Toxicology 325 (2014) 12–20 15

2.7. Methylation-sensitive quantitative PCR (qPCR) analysis of repetitive elements
methylation
The methylation status of long interspersed elements 1 (LINE 1) and short
interspersed nuclear elements (SINE B1 and SINE B2) were determined by
methylation-sensitive McrBC-qPCR assay. Briefly, genomic DNA (1 mg) was
digested overnight with the methylation-specific restriction enzyme McrBC
(New England Biolabs, Ipswich, MA) and then analyzed by qPCR on an Applied
Biosystems (Forrest City, CA) 7900 Real Time PCR System. The threshold cycle (Ct)
was defined as the fractional cycle number that passed the fixed threshold. The
Ct values were converted into the absolute amount of input DNA using the absolute
standard curve method. An increased amount of input DNA after digestion with
McrBC is indicative of hypomethylation, whereas a decreased amount of input DNA
is indicative of hypermethylation.
2.8. Determination of one-carbon metabolites
The levels of S-adenosyl-L-methionine (SAM) and S-adenosyl-L-homocysteine
(SAH) in the livers were measured by HPLC coupled with coulometric
electrochemical detection (Melnyk et al., 2000).
2.9. Western blot analysis of histone modifications
The status of histone H3 lysine 4 (H3K4), histone H3 lysine 9 (H3K9), histone
H3 lysine 27 (H3K27), histone H3 lysine 79 (H3K79), and histone H4 lysine 20
(H4K20) trimethylation and acetylation of histone H3K9, H3K18, H3K27, and H4K16
was determined by Western blotting (Tryndyak et al., 2006).
2.10. RNA extraction and gene expression analysis using microarray technology
Total RNA was extracted from liver tissue using RNeasy Mini kits (Qiagen,
Valencia, CA). Gene expression profiles in the livers of mice fed tamoxifen and
corresponding control mice were determined utilizing Agilent whole genome
4  44 K mouse microarrays (Agilent Technologies, Santa Clara, CA). Sample
labeling and microarray processing were performed as detailed in the “One-Color
Microarray-Based Gene Expression Analysis” Version 5.5 (Agilent Technologies)
protocol. The hybridized slides were scanned with an Agilent DNA Microarray
scanner (Agilent Technologies) at 5 mm resolution. The resulting images were
analyzed by determining the Cy3 fluorescence intensity of all gene spots
(features) on each array using the Agilent Feature Extraction Software (Version
10.7). The raw data were then uploaded into the ArrayTrack database (Fang et al.,
2009). The median fluorescence intensity of all the pixels within one feature was
taken as the intensity value for that feature. The raw intensity values were then
normalized using 75 percentile channel scaling normalization using ArrayTrack. A
list of differentially expressed genes was generated with ArrayTrack using a t-test
at p-value <0.01 and a fold change <2.0.
2.11. Functional analysis of significant genes
Ingenuity Pathway Analysis software (IPA, IPA version 9.0; Ingenuity Systems,
Redwood City, CA) was used to determine canonical pathways that were enriched
for significant mRNA transcripts identified from the t-test analysis. Significance
values were calculated based upon a right-tailed Fisher’s exact test that
determined whether a pathway was overrepresented by calculating whether
the genes in a given pathway were enriched within the data set compared to all
genes on the array; p < 0.05 was selected as the cutoff for significance based on IPA
threshold recommendations. Only those pathways with a p-value below the
threshold and having more than three representative genes in the data set were
considered significant.
2.12. Quantitative reverse-transcription PCR analysis
Total RNA (2 mg) was reverse transcribed using random primers and a High
Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). cDNA
was analyzed in a 96-well plate assay format using a 7900HT Fast Real-Time PCR
System (Applied Biosystems). Each plate contained one experimental gene and a
housekeeping gene. All primers for the gene expression analysis were obtained from
Applied Biosystems. The cycle threshold (Ct) for each sample was determined from the
linear region of an amplification plot. The DCt values for all genes were determined
relative to the endogenous control glyceraldehyde-3-phosphate dehydrogenase
(Gapdh). The DDit values were calculated using treated group means relative to
strain-matched control group means. The fold change data were calculated from the
DDCt values. A gene was considered to be differentially expressed if the change in
expression corresponded to the following criteria: a p-value <0.05 and a fold change
>1.5. All quantitative reverse transcription-PCR (qRT-PCR) reactions were conducted
in triplicate and the experiments were repeated twice.

Fig. 3. DNA methylation, expression of DNA methyltransferases, and hepatic content of S-adenosylmethionine (SAM) and S-adenosylhomocysteine (SAH) in control mice and
mice fed tamoxifen. (A) Global DNA methylation in the livers as measured by a cytosine extension ([3H]dCTP incorporation) DNA methylation assay. (B) Methylation status of
repetitive elements LINE1, SINE B1, SINE B2 in the livers as measured by a methylation-sensitive McrBC-qPCR assay. (C) Expression of DNA methytransferases (Dnmt1,Dnmt3a,
and Dnmt3b) in the livers as measured by qRT-PCR. (D) Hepatic levels of SAM, SAH, and SAM/SAH ratio as detected by HPLC. Values are means  S.D. (n = 4). * – Denotes a
significant (p < 0.05) difference from the control mice.
16 A. de Conti et al. / Toxicology 325 (2014) 12–20

Fig. 4. Histone modifications and expression of chromatin modifying genes in control mice and mice fed tamoxifen. (A) Representative immunoblot images of H3K4me3,
H3K9me3, H3K27me3, H3K79me3, H4K20me3, H3K9ac, H3K18ac, H3K27ac, and H4K16ac in the livers. (B) Quantitative analysis of Western blot images is presented as
percent change in histones methylation or acetylation relative to that of control animals. Equal sample loading was confirmed by immunostaining against histone H3. (C)
Expression of chromatin modifying genes (Suv39h1, Suv420h2, Kdm4a, Kdm4b, Kdm6a, Jhdm1, Prdm2 and Ezh2) in the livers as measured by qRT-PCR. Values are means  S.D.
(n = 5). * – Denotes a significant (p < 0.05) difference from the control mice.

2.13. Statistical analyses
Results are presented as mean  S.D. When appropriate, Student’s t-test or
Mann–Whitney’s test were used. Linear regression analysis was used to determine
time-related trends. When necessary, the data were natural log transformed before
conducting the analyses to maintain a more equal variance or normal data
distribution. p-values <0.05 were considered significant.
changes were found. Tamoxifen did not alter the hepatic levels of
triglycerides and phosphatidylcholine (Fig. 1E). Likewise, the
expression of a-Sma and Col1a1 genes, sensitive indicators of
activation of hepatic fibrogenesis, was not affected by tamoxifen
treatment (Fig. 1F).

3.2. Levels of hepatic tamoxifen-DNA adducts

3. Results

HPLC-ES-MS/MS was used to assess the levels of dG-TAM and

3.1. Morphological and biochemical evaluation in the livers of mice
treated with tamoxifen
Fig. 1 shows that over the course of this 12 week study the body
weight (panel a) and relative liver weight (panel b) of control and
tamoxifen-treated mice did not differ; however, the body weight of
control mice increased in a time-dependent manner, while it did
not change in mice treated with tamoxifen (panel a). The levels of
glucose, triglycerides, total cholesterol, and activity of ALT, AST, and
LDH decreased in plasma of tamoxifen-treated mice (Fig. 1C).
Histomorphological evaluation of liver sections from mice
treated with tamoxifen showed a moderate decrease of glycogen
(Fig. 1D, panels a and b). No differences in the extent of cell
proliferation (Fig. 1D, panels c and d) or other histopathological
dG-DesMeTAM in hepatic DNA of female mice fed tamoxifen and
corresponding control mice (Fig. 2). Table 1 shows that tamoxifen
treatment resulted in a profound formation of total tamoxifen-
DNA adducts, 918 adducts/108 nucleotides, in liver DNA. dG-TAM
and dG-DesMeTAM accounted for 63% and 37% of the total
tamoxifen-DNA adducts, respectively. Neither dG-TAM nor
dG-DesMeTAM (<1 adduct/108 nucleotides) was detected in the
DNA isolated from the livers of control mice.
3.3. Effect of tamoxifen treatment on DNA methylation
Mounting evidence indicates that the presence of DNA adducts
induced by genotoxic carcinogens, including tamoxifen, may

Table 2
Toxicological pathways affected by tamoxifen treatment in the livers of female WSB/EiJ mice.

Toxicity pathway p- Number Gene expression

value of genes
Fatty acid metabolism 0.0001 18 Cyp2d6", Aldh1b1#, Acat2#, Cyp2c40#, Acaa1", Srd5a1", Adh1c#, Cyp2c44#, Acad9#, Cyp2d13#, Aldh9a1#, Cyp2f1#, Cyp2a13/
Cyp2a6#, Cyp3a5#, Cyp4b1", Cyp4f12#, Ehhadh", Acsl1#
Xenobiotic metabolism 0.0002 29 Cyp2d6", Cyp2f1#, Cyp2c40#, Cyp2a13/Cyp2a6#, Cyp3a5#, Cyp2c44#, Cyp2d13#, Cyp3a13#, Lipc#, Ugt1a6", Adh1c#, Sult1a3/
signaling Sult1a4#, Prkcz#, Gstm2#, Nr1i3#, Fmo1#, Fmo4#, Cited2#, Prkca", Aldh1b1#, Fmo3#, Map3k13", Aldh9a1#, Cyp4b1",
Cyp4f12#, Map2k3#, Rxra#, Ndst1#, Gstp1"
Nuclear receptors 0.0417 52 Abcg8", Abcg5", Slc10a1, Lipc#, Cyp2c40#, Sult1a3/Sult1a4#, Cyp2c44#, Gstm2#, Cyp2a13/Cyp2a6#, Cyp3a5#, Ngfr#, Nr1i3#,
signaling Fmo1#, Fmo4#, Cyp2d6", Aldh1b1#, Fmo3#, Tnfrsf1a#, Aldh9a1#, Cyp2f1#, Fabp2#, Rxra#, Ndst1#, Gstp1", Acsl1#, Pparg",
Pon1#, Pklr#, Fgfr4#, Foxa3#, Cyp8b1#, Nr2f1", Stat5a", Nfkbia", Hspa1a/Hspa1b", Lpl", Ehhadh", Cited2#, Prkca", Aldh1b1#,
Cyp2c40#, Cyp2c44#, Rarg", Aldh9a1#, Cyp3a13#, Rxrg#, Gstm2#, Cyp3a5#, E2f1#, Nfib", Esr1#, Gstp1"
Hepatic cholestasis 0.0263 14 Abcg8", Abcg5", Slc10a1#, Tnfrsf1a#, Tap1#, Prkcz#, Atp8b1", Cyp8b1#, Nfkbia", Fgfr4#, Ngfr#, Rxra#, Esr1#, Prkca"
 A. de Conti et al. / Toxicology 325 (2014) 12–20 17

Fig. 5. Expression of selective lipid metabolism genes in the livers of control mice and mice treated with tamoxifen. (A) Molecular network interactions among genes in the
livers of WSB/EiJ mice significantly differ between control and tamoxifen-treated groups. The Ingenuity Pathway Analysis software (version 9.0) was used to determine and
visualize molecular pathways enriched by the significant mRNA transcripts. Up-regulated genes are marked in red and down-regulated genes in green. (B) The expression of
Acox1, Mcad, Fasn, Ppara, Pparg, Pparyc1a, and Cebpa genes was determined by qRT-PCR as detailed in Section 2. The levels of CEBPa and PPARa proteins were determined by
Western blotting. The results are presented as an average fold change in the livers of tamoxifen-treated mice relatively to that in control groups, which were assigned a value 1.
Values are means  S.D. (n = 5). * – Denotes a significant (p < 0.05) difference from the control mice.

perturb the cellular epigenomic status by causing a range of aberrations. In view of this, the status of global DNA methylation
epigenetic alterations (Chen et al., 2004; FitzGerald and Shank, and methylation of repetitive elements in the livers of control mice
1996; Tryndyak et al., 2006), among which demethylation of the and mice exposed to tamoxifen was examined. The extent of DNA
genome is one of the most prominent carcinogen-induced methylation (Fig. 3A) and methylation of repetitive elements
18 A. de Conti et al. / Toxicology 325 (2014) 12–20
LINE1, SINE B1 and SINE B2 (Fig. 3B) in the livers of mice fed
tamoxifen did not differ from that in control mice.
It has been established that proper DNA methylation depends,
in addition to DNA integrity, on the function of DNA methyl-
transferases (DNMTs) and status of cellular one-carbon metabo-
lism (Huidobro et al., 2013). Hence, the expression of Dnmt genes
and the levels of SAM and SAH, key indicators of DNA methylation
capacity, in the livers of mice treated with tamoxifen were
investigated. Fig. 3C shows that the expression of de novo DNA
methyltransferases Dnmt3a and, especially, Dnmt3b was reduced
in the livers of tamoxifen-treated mice, while the expression of
Dnmt1, a main DNA methyltransferase in somatic mammalian cells,
was not affected. Also, treatment with tamoxifen did not alter the
hepatic level of SAM, SAH, and SAM/SAH ratio (Fig. 3D).
3.4. Effect of tamoxifen treatment on histone modifications
Genotoxic carcinogens may also compromise the status of
histone modifications in the target organ (Bagnyukova et al., 2008;
Huidobro et al., 2013). Therefore, the effect of tamoxifen on the
status of selected histone modifications that play a major role in
the chromatin architecture and gene expression in the livers was
examined. Tamoxifen treatment increased slightly the levels of
hepatic histone H3K4me3, while no changes were found in
methylation of histone H3K9, H3K27, H3K79, and H4K20, and
acetylation of histone H3K9, H3K18, H3K27, and H4K16 (Fig. 4A
and B). Fig. 4C shows that the expression of chromatin modifying
genes Suv39h1, Kdm4b, Kdm6a, Jhdm1d, Prdm2, and Ezh2 was not
affected by tamoxifen treatment. Only expression of histone
H4K20 methyltransferase Suv420h2 was significantly reduced in
the livers of tamoxifen-treated mice.
3.5. Effect of tamoxifen on hepatic gene expression profile
To investigate further the effect of tamoxifen on molecular
processes in the livers, hepatic gene expression profiles were
examined in control and tamoxifen-treated mice using high-
throughput Agilent whole genome 4  44 K mouse microarrays. To
identify genes that were differentially expressed between the
control and tamoxifen groups, a t-test, p < 0.01, coupled with a
fold-change cut-off >2 was applied. A total of 462 significantly
up-regulated or down-regulated genes was identified in the livers
of mice treated with tamoxifen as compared to control mice
(Supplementary Table 1). Pathway enrichment analysis of these
differentially expressed hepatic genes demonstrated a strong
enrichment in genes associated with the xenobiotic metabolism,
fatty acid metabolism, and nuclear receptors signaling (Table 2).
Treatment with tamoxifen increased the expression of Cyp2d6,
which is one of the most active enzymes responsible for the
metabolic activation of tamoxifen and formation of
4-hydroxytamoxifen (Boocock et al., 2002). The expression of this
gene in the livers of tamoxifen-treated mice was 2.4-times greater
than in the control group. In contrast, the expression of other
tamoxifen-metabolizing cytochrome P450 and flavin-containing
monooxygenase genes, especially Fmo3, which catalyzes
N-oxidation of tamoxifen, was reduced. Specifically, the expression
Fmo3 was 41.6 times lower than in the livers of control mice
(Supplementary Table 1).
Interestingly, the greatest effect of tamoxifen was on genes
involved in lipid metabolism pathways (Table 2; p < 0.0001).
Specifically, treatment with tamoxifen caused predominantly
down-regulation of hepatic lipid metabolism genes (Fig. 5).
However, one of the most noticeable changes induced by
tamoxifen was a distinct up-regulation of the lipocalin 13
(Lcn13) and peroxisome proliferator receptor gamma (Pparg )
genes. The Lcn13 gene was the most up-regulated gene
(Supplementary Table 1), and the expression of Pparg in the
livers of tamoxifen-treated mice was 10-times greater as compared
to control mice (Fig. 5B).
4. Discussion
Genotoxic carcinogens are agents that interact with DNA, either
directly or after metabolic activation, leading to the formation of
DNA adducts, which is the indispensable event in cancer initiation.
However, it is well-established even within classical genotoxic
carcinogenesis models that the formation of DNA adducts and/or
damage to DNA is necessary but not sufficient for tumor formation.
In the present study, we demonstrate that exposure of female mice
to tamoxifen for 12 weeks resulted in the formation of tamoxifen-
DNA adducts in the livers. Interestingly, the levels of hepatic
tamoxifen DNA-adducts in mice were similar to those found in the
livers of rats exposed to the same dose of tamoxifen for the same
period of time (Tryndyak et al., 2006). This is contrary to the report
by Umemoto et al. (2001) showing a lower level of tamoxifen-DNA
adduct formation in mouse liver. The results of this study
demonstrated that dG-TAM adducts were formed in the livers of
mice to a greater extent than dG-desMeTAM adducts as compared
to those observed in the livers of tamoxifen-treated rats (Tryndyak
et al., 2006). This may be due to the induction of Cyp3A isoforms in
the rat (Pogribny et al., 2007), which are responsible for the N-
demethylation of tamoxifen. This finding is in good agreement
with a suggestion that the quantity of DNA adducts formed by a
DNA-reactive compound is not predictive of carcinogenicity across
tissues and species (Paini et al., 2011). It also emphasizes the need
of better understanding other factors and mechanisms that may be
equally relevant to the carcinogenic process.
Over the past decade, it has become clear that epigenetic
changes, such as DNA methylation and histones modifications, are
important in carcinogenesis as genetic alterations (Baylin and
Jones, 2012). Previously, we demonstrated that hepatocarcino-
genic activity of tamoxifen in female rats is associated with its
ability to induce, in addition to genotoxic alterations, extensive
epigenetic abnormalities. Specifically, tamoxifen treatment has led
to early extensive cancer-related epigenetic changes in rat liver
tissue, including global DNA and repetitive elements demethyla-
tion, and progressive loss of histone H4K20 trimethylation
(Tryndyak et al., 2006, 2007). In contrast to those results, in the
present study treatment with tamoxifen for 12 weeks did not
induce changes in the status of DNA methylation or post-
translational histone modifications, except for a moderate increase
in histone H3K4 trimethylation, in the mouse liver.
The resistance of mice to tamoxifen-induced liver carcinogen-
esis may be explained, in part, by the induction of Cyp2d6 and an
inability of tamoxifen to cause epigenetic dysregulation in the liver.
Specifically, the induction of Cyp2d6 in mice might result in a
greater formation of 4-hydroxytamoxifen that binds to the
estrogen receptor and prevents estrogen-associated hepatocarci-
nogenesis. The results of the present study suggest that the
resistance of mice to tamoxifen-induced liver carcinogenesis may
be associated with an inability of tamoxifen to provoke non-
genotoxic events in the liver, which are critical to the hepato-
carcinogenic process.
It has been reported that tamoxifen may cause the development
of NAFLD in humans (Bruno et al., 2005; Murata et al., 2000; Oien
et al., 1999; Saphner et al., 2009) and experimental animals
(Cole et al., 2012; Lelliott et al., 2005). However, there is a lack of
conclusive information on the mechanism of the tamoxifen-induced
NAFLD. For instance, Cole et al. (2012) have reported that induction of
hepatic steatosis in male C57BL/6J mice in a short-term tamoxifen
treatment, 0.5 mg/kg body weight for 5 consecutive days, was linked
to an increased de novo fatty acid synthesis. A similar treatment of
 A. de Conti et al. / Toxicology 325 (2014) 12–20
male Wistar rats also caused steatosis; however, the mechanism of
steatosis induction was through inhibition of fatty acid synthesis
(Lelliott et al., 2005), an opposite mechanism to that in the mouse
model. In contrast, in the present study we did not find any
biochemical or histopathological evidence of NAFLD-associated liver
injury in female WSB/EiJ mice treated with tamoxifen for 12 weeks.
The discrepancy of our findings with the results of those two studies
may be caused by species, gender, and mouse strain difference.
Furthermore, to a report by Miyashita et al. (2012) showing that
tamoxifen treatment exerts a hepatoprotective effect against
steatosis in mouse-models of NAFLD supports our findings. This
suggests that another potential mechanism of the resistance of mice
to tamoxifen-induced hepatocarcinogenesis may be attributed also
to its lack of ability to induce the development of NAFLD.
A transcriptomic analysis of differentially expressed genes
identified four major pathways, including tamoxifen metabolism
and detoxification, fatty acid metabolism, and nuclear receptor
signaling, that were affected by tamoxifen treatment. The results of
the gene expression analysis demonstrate that tamoxifen treat-
ment caused inhibition of de novo fatty acid synthesis. Similar
changes were found in the livers of tamoxifen-treated rats
(Pogribny et al., 2007); however, in the present study we found
a marked increase in the expression of Lcn13, which was the most
up-regulated gene in the livers of tamoxifen-treated mice. It has
been demonstrated that Lcn13 inhibits the expression of lipogenic
genes and stimulates genes that promote b-oxidation and prevent
the development of NAFLD (Sheng et al., 2011).
Another important finding of the present study is a marked
increase in the expression of Pparg. In our previous study on mouse
interstrain differences in the severity of NAFLD caused by feeding a
choline- and folate deficient diet, we demonstrated that the A/J
mice, most resistant strain to diet-induced NAFLD-associated liver
injury, exhibited a similar gene expression pattern–a moderate
down-regulation of lipid catabolism genes, and up-regulation of
PPARg (Tryndyak et al., 2012). Importantly, the extent of Pparg
over-expression observed in the present study was substantially
greater than in the resistant A/J mice in our previous study. This
suggests that up-regulation of PPARg-regulated cellular networks
may be one of the mechanisms protecting WSB/EiJ mice from the
development of NAFLD-associated liver injury. This suggestion is
supported by a report of Nan et al. (2011) demonstrating that
up-regulation of PPARg amended NAFLD-associated features, such
as hepatic steatosis, necroinflammation, and fibrosis, induced by
methionine- and choline deficient diet in C57BL/6J mice. Addi-
tionally, mounting evidence indicates that stimulation of PPARg
regulatory system has positive therapeutic effect on main
constituents of metabolic syndrome: obesity, insulin resistance,
and hepatic steatosis and fibrosis (Ables, 2012; Derosa and Maffioli,
2012; Zhang et al., 2013; Zheng et al., 2011).
In conclusion, the results of the present study demonstrate that
the resistance of mice to tamoxifen-induced liver carcinogenesis is
associated with its ability to induce genotoxic alterations only
without affecting the cellular epigenome. These results provide
additional evidence that carcinogenic process requires both
genotoxic and non-genotoxic alterations. Additionally, these data
reinforce the importance of epigenetic alterations as indispensable
events in the carcinogenic process.
Conflict of interest
The authors declare that there are no conflicts of interest.
Transparency document
The Transparency document associated with this article can be
found in the online version.
19
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tox.2014.08.004.
References
Ables, G.P., 2012. Update on Pparg and nonalcoholic fatty liver disease. PPAR Res.
912351.
Bagnyukova, T.V., Tryndyak, V.P., Montgomery, B., Churchwell, M.I., Karpf, A.R.,
James, S.R., Muskhelishvili, L., Beland, F.A., Pogribny, I.P., 2008. Genetic and
epigenetic changes in rat preneoplastic liver tissue induced by
2-acetylaminofluorene. Carcinogenesis 29, 638–646.
Baylin, S.B., Jones, P.A., 2012. A decade of exploring the cancer epigenome–biological
and translational implications. Nat. Rev. Cancer 11, 726–734.
Boocock, D.J., Brown, K., Gibbs, A.H., Sanchez, E., Turteltaub, K.W., White, I.N.H.,
2002. Identification of human CYP forms involved in the activation of tamoxifen
and irreversible binding to DNA. Carcinogenesis 23, 1897–1901.
Bruno, S., Maisonneuve, P., Castellana, P., Rotmensz, N., Rossi, S., Maggioni, M.,
Persico, M., Colombo, A., Monasterolo, F., Casadei-Giunchi, D., Desiderio, F.,
Stroffolini, T., Sacchini, V., Decensi, A., Veronesi, U., 2005. Incidence and risk
factors for non-alcoholic steatohepatitis: prospective study of 5408 women
enrolled in Italian tamoxifen chemoprevention trial. BMJ 330, 932.
Chen, H., Li, S., Liu, J., Diwan, B.A., Barret, J.C., Waalkes, M.P., 2004. Chronic inorganic
arsenic exposure induces hepatic global and individual gene hypomethylation:
implications for arsenic hepatocarcinogenesis. Carcinogenesis 25, 1779–1786.
Cole, L.K., Jacobs, R.L., Vance, D.E., 2012. Tamoxifen induces triacylglycerol
accumulation in the mouse liver by activation of fatty acid synthesis.
Hepatology 52, 1258–1265.
Davies, R., Oreffo, V.I.C., Martin, E.A., Festing, M.F.W., White, I.N.H., Smith, L.L., Styles,
J.A., 1997. Tamoxifen causes gene mutations in the livers of Lambda/lacI
transgenic rats. Cancer Res. 57, 1288–1293.
Derosa, G., Maffioli, P., 2012. Peroxisome proliferator-activated receptor-(PPAR-g)
agonists on glycemic control, lipid profile and cardiovascular risk. Curr. Mol.
Pharmacol. 5, 272–281.
Fang, H., Harris, S.C., Su, Z., Chen, M., Qian, F., Shi, L., Perkins, R., Tong, W., 2009.
ArrayTrack: an FDA and public genomic tool. Methods Mol. Biol. 563, 379–398.
Fisher, B., Costantino, J.P., Redmond, C.K., Fisher, E.R., Wickerham, D.L., Cronin, W.M.,
1994. Endometrial cancer in tamoxifen-treated breast cancer patients: findings
from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14. J.
Natl. Cancer Inst. 86, 527–537.
FitzGerald, B.E., Shank, R.C., 1996. Methylation status of DNA cytosine during the
course of induction of liver cancer in hamsters by hydrazine sulfate.
Carcinogenesis 17, 2703–2709.
Gamboa da Costa, G., Marques, M.M., Beland, F.A., Freeman, J.P., Churchwell, M.I.,
Doerge, D.R., 2003. Quantification of tamoxifen–DNA adducts using on-line
sample preparation and HPLC-electrospray ionization tandem mass spectrom-
etry. Chem. Res. Toxicol. 16, 357–366.
Greaves, P., Goonetilleke, R., Nunn, G., Topham, J., Orton, T., 1993. Two-year
carcinogenicity study of tamoxifen in Alderley Park Wistar-derived rats. Cancer
Res. 53, 3919–3924.
Huidobro, C., Fernandez, A.F., Fraga, M.F., 2013. The role of genetics in the
establishment and maintence of the epigenome. Cell. Mol. Life Sci. 70,
1543–1573.
Jones, M.E., van Leeuwen, F.E., Hoogendooorn, W.E., Mourits, M.J., Hollema, H., van
Boven, H., Press, M.F., Bernstein, L., Swerdlow, A.J., 2012. Endometrial cancer
survival after breast cancer in relation to tamoxifen treatment: pooled results
from three countries. Breast Cancer Res. 14, R91.
Kiyotani, K., Mushiroda, T., Nakamura, Y., Zembutsu, H., 2012. Pharmacogenomics of
tamoxifen: roles of drug metabolizing enzymes and transporters. Drug Metab.
Pharmacokinet. 27, 122–131.
Lelliott, C.J., Lopez, M., Curtis, R.K., Parker, N., Laudes, M., Yeo, G., Jimenez-Linan, M.,
Grosse, J., Saha, A.K., Wiggins, D., Hauton, D., Brand, M.D., O'Rahilly, S., Griffin, J.
L., Gibbons, G.F., Vidal-Puig, A., 2005. Transcript and metabolite analysis of the
effects of tamoxifen in rat liver reveals inhibition of fatty acid synthesis in the
presence of hepatic steatosis. FASEB J. 19, 1108–1119.
Martin, E.A., Carthew, P., White, I.N.H., Heydon, R.T., Gaskell, M., Mauthe, R.J.,
Turteltaub, K.W., Smith, L.L., 1997. Investigation of the formation and
accumulation of liver DNA adducts in mice chronically exposed to tamoxifen.
Carcinogenesis. 18, 2209–2215.
Melnyk, S., Pogribna, M., Pogribny, I.P., Yi, P., James, S.J., 2000. Measurement of
plasma and intracellular S-adenosylmethionine and S-adenosylhomocysteine
utilizing coulometric electrochemical detection: alterations with plasma
homocysteine and pyridoxal 50 -phosphate concentrations. Clin. Chem. 46,
265–272.
Miyashita, T., Toyoda, Y., Tsuneyama, K., Fukami, T., Nakajima, M., Yokoi, T., 2012.
Hepatoprotective effect of tamoxifen on steatosis and non-alcoholic steatohe-
patitis in mouse models. J. Toxicol. Sci. 37, 931–942.
Murata, Y., Ogawa, Y., Saibara, T., Nishioka, A., Fujiwara, Y., Fukumoto, M., Inomata,
T., Enzan, H., Onishi, S., Yoshida, S., 2000. Unrecognized hepatic steatosis and
non-alcoholic steatohepatitis in adjuvant tamoxifen for breast cancer patients.
Oncol. Rep. 7, 1299–1304.
Nan, Y.-M., Han, F., Kong, L.-B., Zhao, S.-X., Wang, R.-Q., Wu, W.-J., Yu, J., 2011.
Adenovirus-mediated peroxisome proliferator activated receptor gamma
20 A. de Conti et al. / Toxicology 325 (2014) 12–20
overexpression prevents nutritional fibrotic steatohepatitis in mice. Scand. J.
Gastroenterol. 46, 358–369.
Oien, K.A., Moffat, D., Curry, G.W., Dickson, J., Habeshaw, T., Mills, P.R., MacSween, R.N.,
1999. Cirrhosis with steatohepatitis after adjuvant tamoxifen. Lancet 353,
36–37.
Paini, A., Scholz, G., Marin-Kuan, M., Schilter, B., O'Brien, J., van Bladeren, P.J.,
Rietjens, I.M.C.M., 2011. Quantitative comparison between in vivo DNA adduct
formation from exposure to selected DNA-reactive carcinogens, natural
background levels of DNA adduct formation and tumour incidence in rodent
bioassays. Mutagenesis 26, 605–618.
Pogribny, I., Yi, P., James, S.J., 1999. A sensitive new method for rapid detection of
abnormal methylation patterns in global DNA and within CpG islands. Biochem.
Biophys. Res. Commun. 262, 624–628.
Pogribny, I.P., Bagnyukova, T.V., Tryndyak, V.P., Muskhelishvili, L., Rodriguez-Juarez,
R., Kovalchuk, O., Han, T., Fuscoe, J.C., Ross, S.A., Beland, F.A., 2007. Gene
expression profiling reveals underlying molecular mechanisms of the early
stages of tamoxifen-induced rat hepatocarcinogenesis. Toxicol. Appl. Pharma-
col. 225, 61–69.
Saphner, T., Triest-Robertson, S., Li, H., Holzman, P., 2009. The association of
nonalcoholic steatohepatitis and tamoxifen in patients with breast cancer.
Cancer 115, 3189–3195.
Sheng, L., Cho, K.W., Zhou, Y., Shen, H., Rui, L., 2011. Lipocalin 13 protein proitects
against hepatic steatosis by both inhibiting lipogenesis and stimulating fatty
acid b-oxidation. J. Biol. Chem. 286, 38128–38135.
Tryndyak, V.P., Muskhelishvili, L., Kovalchuk, O., Rodriguez-Juarez, R., Montgomery,
B., Churchwell, M.I., Ross, S.A., Beland, F.A., Pogribny, I.P., 2006. Effect of long-
term tamoxifen exposure on genotoxic and epigenetic changes in rat liver:
implications for tamoxifen-induced hepatocarcinogenesis. Carcinogenesis 27,
1713–1720.
Tryndyak, V.P., Kovalchuk, O., Muskhelishvili, L., Montgomery, B., Rodriguez-Juarez,
R., Melnyk, S., Ross, S.A., Beland, F.A., Pogribny, I.P., 2007. Epigenetic
reprogramming of liver cells in tamoxifen-induced rat hepatocarcinogenesis.
Mol. Carcinog. 46, 187–197.
Tryndyak, V., de Conti, A., Kobets, T., Kutanzi, K., Koturbash, I., Han, T., Fuscoe, J.C.,
Latendresse, J.R., Melnyk, S., Shymonyak, S., Collins, L., Ross, S.A., Rusyn, I., Beland, F.
A., Pogribny, I.P., 2012. Interstrain differences in the severity of liver injury induced
by a choline- and folate-deficient diet in mice are associated with dysregulation of
genes involved in lipid metabolism. FASEB J. 26, 4592–4602.
Tucker, M.J., Adams, H.K., Patterson, J.S., 1984. Tamoxifen. In: aurence, D.R.,
McLean, A.E.M., Weatherall, M. (Eds.), Safety Testing of New Drugs. Academic
Press, New York, pp. 125–161.
Umemoto, A., Komaki, K., Monden, Y., Suwa, M., Kanno, Y., Kitagawa, M., Suzuki, M.,
Lin, C.-X., Ueyama, Y., Momen, M.A., Ravindernath, A., Shibutani, S., 2001.
Identification and quantification of tamoxifen-DNA adducts in the liver of rats
and mice. Chem. Res. Toxicol. 14, 1006–1013.
Zhang, F., Kong, D., Lu, Y., Zheng, S., 2013. Peroxisome proliferator-activated
receptor-g as a therapeutic target for hepatic fibrosis: from bench to beside.
Cell. Mol. Life Sci. 70, 259–276.
Zheng, H., Li, S., Ma, L., Cheng, L., Deng, C., Chen, Z., Xie, C., Xiang, M., Jiang, W., Chen,
L., 2011. A novel agonist of PPAR- g based on barbituric acid alleviates the
development of non-alcoholic fatty liver disease by regulating adipocytocine
expression and preventing insulin resistance. Eur. J. Pharmacol. 659, 244–251.