Acute exposure to DEHP metabolite, MEHP cause genotoxicity, mutagenesis and carcinogenicity in mammalian Chinese hamster ovary cells
Acute exposure to DEHP metabolite, MEHP cause genotoxicity, mutagenesis and carcinogenicity in mammalian Chinese hamster ovary cells
Yu-Jung Chang 2
Chia-Yi Tseng 0 1
Pei-Ying Lin 2
Yu-Chen Chuang 1
Ming-Wei Chao 0 2
0 Center of Nanotechnology, Chung Yuan Christian University , Taoyuan 32023 , Taiwan
1 eD, epartment of Biomedical Engineering, College of Engineering
2 Department of Bioscience Technology, College of Scienc
Di-(2-ethylhexyl) phthalate (DEHP), the common plasticizer used in the production of polyvinyl chloride, can be converted to the more potent metabolite mono-ethylhexyl phthalate (MEHP). Epidemiological studies have shown an association with elevated induction of rat hepatic cancer and reproductive toxicity in response to MEHP exposure. However, the mechanism of genotoxicity and carcinogenicity induced by MEHP treatment remains unclear. As a means to elucidate the mechanisms of action, lethality and mutagenicity in the adenine phosphoribosyltransferaaspert(+/−) gene induced in several CHO cell types by MEHP were assessed. Dose-response relationships were determined in the parental AA8 cell line, its nucleotide repair-deficient UV5 and base repair-deficient EM9 subclones, and also in AS52 cells harboring the bacterial guaninehypoxanthine phosphoribosyltransferaseg(pt) gene and its derived AS52-XPD-knockdown and AS52-PARP-1-knockdown cells. Treatment of AS52 with MEHP led to intracellular production of reactive oxygen species (ROS) and DNA strand breaks in a dose-dependent manner. Separately, mutations in thgept gene of AS52 cells were characterized and found to be dominated by G:C to A:T and A:T to G:C transitions. Independent AS52-mutant cell (ASMC) clones were collected for the sequentialin vivo xenograft tumorigenic studies, 4 of total 20 clones had aggressive tumor growth. Moreover, microarray analysis indicated miR-let-7a and miR-125b downregulated in ASMC, which might raise oncogenMicYC and RAS level and activate ErbB pathway. Comparative evaluation of the results indicates that the principal mechanism of this mutagenic action is probably to be through generation of ROS, causing base excision damage resulting in carcinogenicity.
Di-(2-ethylhexyl) phthalate (DEHP) is a common industrial has indicated that MEHP may cause toxicity and carcinogenicity
chemical substance. DEHP is able to extend the ductility and in rodent hepatocytes through PPARα-(4,5) and may also induce
malleability of plastic and is widely used in polyvinyl chloride oxidative stress to other cells and cause DNA damage and ap-op
plastic products such as medical supplies, plastic bags and tosis (6,7).
toys (1). Previous research showed that the intestinal hyd-ro DNA is sensitive to environmental factors. Such factors can
lytic enzymes of mammals can transform DEHP into its primary affect DNA stability and are probably to damage genes. Damage
metabolite mono-ethylhexyl phthalate (MEHP), which has a to DNA will cause cells to initiate DNA repair using the base
higher toxicity than DEHP and has both reproductive and dev-el excision repair mechanism to remove lesioned DNA bases (8),
opmental toxicity 2( ,3). Though the carcinogenic mechanisms of the nucleotide excision pathway to repair bulky gene damage, or
both DEHP and MEHP still remain to be clarified, prior research the mismatch repair system (9). If these damages to DNA are not
AS52 mutant cells
base excision repair
hypoxanthine phosphoribosyl transferase
reactive oxygen species
small interfering RNA
Cell survival in AA8, UV5 and EM9 cells with and without N-acetyl cysteine
AA8, UV5, EM9 and AS52 cells were plated at ×8 103 cells/well in
96-well plates and incubated overnight in medium prior to the
treatment. After 2 h MEHP treatment, cells were washed with
phosphate-buffered saline (PBS) and incubated in medium for
additional 0 and 24 h prior to determining cell survival. Viability
was detected using a CellTiter 96® AQueous One Solution Cell
Proliferation Assay (MTS assay) (Promega, USA). The absor-b
ance of the generated formazan was measured at 490 nm with
enzyme-linked immunosorbent assay (ELISA) reader (Promega,
RNA interference and transfection
completely or correctly repaired, there is an increased chance Small interfering RNA (siRNA) for PARP-1 (sense ′5-GACCUA
for gene mutations, and cancer may be one of the induced AAGGAGUUGCUUA-3′, antisence 5′-UAAGCAACUCCUUU
downstream diseases. AGGUC-3′) and XPD (sense 5′-CCAUUAUCAUCGAGCCCUU-3′,
Previous research involved a model that consists of CHO cells antisence 5′-AAGGGCUCGAUGAUAAUGG-3′) and control
that have a mutation in the purine salvage pathway enzyme siRNA (sense 5′-GAUCAUACGUGCGAUCAGA-3′, antisence
hypoxanthine phosphoribosyl transferase (HPRT). Treatment of 5′-UCUGAUCGCACGUAUGAUC-3′) were designed and
synthethese cells with 6-thioguanine (6-TG) has been used to select sized by Sigma–Aldrich. Cells were transfected with 30 nM
HPRT-mutated cells that have 6-TG resistance and thereby siRNA with Lipofectamine 3000 (Invitrogen, USA) according to
measure the mutation rate of mutagens1(0–12). Tindallet al. the manufacturer’s protocol. The transfected cells were -col
used a pSV2gpt plasmid to transfer thegpt gene of Escherichia lected and performed in following experiment.
coli into the HPRT-deficient CHO cells in order to produce CHO
AS52 cells, which have theE.coli gpt gene. Relative to HPRT-CHO Quantification of intracellular ROS and O-2 generation
cells, the same mutagen will have a higher mutation rate with Intracellular ROS detection studies were using a C2mDHCFDA
AS52 cells 1(3). As a result, AS52 cells are being used in many ROS detection kit (Invitrogen, USA). After 2 h MEHP treatment,
studies as the model for detection of the mutagenicity of toxic AS52 cells were washed with PBS, treated with trypsin-ethy-len
chemicals (14–16). ediaminetetraacetic acid for 5 min and suspended in 0.2 ml/well
Phillips et al. showed that MEHP has mutagenicity in the medium. Aliquots of 100 μl cell suspension from each dose were
HPRT-CHO model. In this study, we used AS52 cells to evaluate pipetted into 96-well black plates and mixed with 10μ0l Hanks’
whether MEHP is mutagenic and oncogenic as a result of its balanced salt solution containing Cm2DHCFDA (final concentr-a
genotoxicity. The results show that acute exposure to MEHP tion 10 μM), incubated for 10 min, at 37°C in the dark. ROS ge-n
produces endogenic oxidative stress in AS52 cells, causing eration was measured at 510 nm by ELISA reader. The ROS levels
DNA base excision damage, which leads to the occurrence of was normalized and presented as the fold change of the control.
point mutations due to DNA mismatch repair. We implanted IntracellulaOr-2 was detected by MitoSOXTM Red mitochondrial
the mutated cells into nude mice and found that the mutated superoxide indicator (Invitrogen, USA). The treated AS52 cells
cells induced by MEHP are carcinogenic. The present study were washed with Hanks’ balanced salt solution and cultured
that we elucidated the role of MEHP-induced reactive o-xy with 5 μM reagent for 10 min incubation, at 37°C in the dark.
gen species (ROS) for genotoxicity, providing the evidence for Samples were measured at 510 nm by ELISA reader. The levels
the carcinogenicity of metabolite MEHP in mammalian cell, of O-2 was normalized and presented as the fold change of the
as well as presenting insight into epigentic modification for control.
Determination of lipid peroxidation
AS52 cells were plated at ×5 105 cells/well in 6-well plate. After
Materials and methods 2 h MEHP treatment, malondialdehyde levels were assessed by
using the lipid peroxidation (malondialdehyde) assay (Sigma–
Cell culture Aldrich) according to manufacturer’s instructions. The levels
Chinese hamster ovary AA8, UV5 and EM9 cells were pu-r of lipid peroxidation was normalized and presented as the fold
chased from Bioresource Collection and Research Center change of the control.
(BCRC, Hsinchu, Taiwan). The three cell lines were derived from
American Type Culture Collection according to manufacturer’s Comet assay
instructions. Cells were cultured in modified Eagle’s medium The alkaline comet assay, used to detect single-strand DNA
culture media (Sigma–Aldrich, St. Louis, MO) with 10% heat- breaks. The protocol described by Woodet al. (2010). AS52 cells
inactivated fetal bovine serum (Gibco). CHO AS52 cells, kindly were placed in 6-well plates and exposed to MEHP ±N-acetyl
provided by Dr G.Wogan (Massachusetts Institute of Technology, cysteine (NAC). After treatment, 5μ0l of cells (150cells/ml) were
USA). AS52 cells were used as model in previous research pipetted into each of the agarose. The slide was covered with
(13–16). Cells were cultured in Ham’s F12 media (Sigma–Aldrich) 1% low melting point agarose. Comet slides were assessed by
with 10% fetal bovine serum. All the cell lines were cultured in a using CometAssay® Electrophoresis Starter Kit (Trevigen, USA)
humidified atmosphere of 5% CO 2 at 37°C. The regular medium according to the manufacturer’s instructions. Lastly, the Comet
was changed routinely, and all cell lines were subcultured when slides were stained with SYB®R Green (Sigma–Aldrich) for the
confluence reached ~80%. fluorescence imaging. Images were captured using an Olympus
IX51 fluorescent microscope coupled with an automatic sca-n ning stage and analyzed using software Image J.
Gpt mutagenesis in CHO AS52 cells
invaded to the lower chamber. The ratio presented as the fold
change of normal AS52 cells. Anchorage-independent growth
was tested by embedding AS52 mutant cells (ASMC, 2.5× 103)
in a 0.35% agarose (Sigma–Aldrich)/growth medium gel overlaid
on solidified 0.5% agarose-containing medium in six-well plate.
After 14 days incubation, these colonies were fixed with 4% PFA
and stained with crystal violet.
The detail method of cleaned of pre-existinggpt mutant AS52
cells were described in previous study Chaoet al. (2012). After
mutant cleaning treatment, cells were plated at× 51 05 cells/
well in 6-well plate. Cells were exposed to MEHP ± NAC for 2 h.
After the treatment, cells were washed with PBS and placed Xenograft animal model
in the regular medium for additional 7 days culture. After All experiments were conducted in accordance with the pr-in
the incubation, 5 × 105 cells from each group were placed in ciples of animal care and experimentation in the Guide for the
100 ml regular Ham’s medium containing selection agent 6-TG Care and Use of Laboratory Animals. The Institutional Animal
(10 μM) and plated at 5× 104 cells/10 ml/100-mm dish. After Care and Use Committee of the Chung Yuan Christian University
14 days incubation, colonies were stained with 0.5% crystal approved the use of animals in this study (IACUC ID#: 10105).
violet (Sigma–Aldrich), rinsed, and counted. The spontaneous The 4 week old female nude mice were obtained from the LASCO
mutant frequency was determined using the negative control. Laboratory (Taipei, Taiwan) in specific pathogen-free conditions.
In addition, twenty-five hundred cells were suspended in 50 ml Mice were obtained excellent care with standard 12/12 light/
medium with 10% fetal bovine serum from each dose which dark cycle with temperature at 18–26°C and humidity at 30–70%.
were seeded in 100-mm dishes at a density of 500 cells/10 ml for Twenty 6TGr subclonal ASMC lines were randomly chosen and
plating efficiency control. incubated in regular medium with 6-TG (10μ M) for 10
generations. 2 × 106 cells in 0.2 ml of PBS were subcutaneously injected
DNA extraction, PCR amplification and molecular into the dorsal flank of each mouse. After tumor formation, the
analysis of gpt mutants mice were sacrificed when the tumor size reached 2 cm.
Single unstained gpt mutant colonies were identified and
transferred to 100 mm dishes and grown to ~5× 106 mutant Hematoxylin and eosin and immunofluorescence
cells. Genomic DNA was extracted from each mutant using stain
GenEluteTM mammalian genomic DNA miniprep kit (Sigma– Tumors were perfused and fixed in 4% PFA. The tumors were
Aldrich). Amplification of the genomic DNA was performed in embedded in FSC 22 Frozen section media (Leica, Germany) and
two rounds of nested PCR in a VeritiTM 96-Well Thermal Cycler soaked in liquid nitrogen. Frozen sections were prepared atμ4m
(Applied Biosystems, USA). Twenty nanogram of template DNA thickness and fixed by 4% PFA. The sections were sequentially
was used to run the first round with 1μ0l Taq DNA polymer- prepared with the hematoxylin and eosin staining protocol.
ase, 8 μl sterile water, 1μ l each 3.2 μM forward (bases −199 to For immunofluorescence staining, the sections were heating
−181; 5′-AAGCTTGGACACAAGACAG-3′), reverse (bases 520–540; for antigen retrieval and then blocked with bovine serum a-lbu
5′-CCAGAATACTTACTGGAAAC-3′) primers and amplified with min blocking buffer. The sections were incubated with mon-o
a PCR profile of 95°C: 5 min, 30 cycles of 95°C: 0.5 min, 47°C: clonal anti-rabbict-myc (1:200, Abcam). Then, the sections were
0.5 min, 72°C: 0.5 min and a final extension of 72°C for 10 min. incubated with second Alexa 488-conjugated goat anti-rabbit
The product from this reaction was filtered using a EasyPure IgG (Jackson ImmunoReasearch, USA) for 1 h. The sections
PCR/Gel Extraction Kit (Bioman Scientific CO., LTD, Taiwan) and were mounted with Fluoromount-G™ (eBioscience, USA)
DAPIsuspended in 100 μl sterile water to avoid non-specific binding Fluoromount-G and observed on an Olympus IX51 fluorescent
with remaining primers, and a 10 μl aliquot was used as tem- microscope.
plate in the second round of PCR using nested primers (bases
−23 to −4; 5′-ATAAACAGGCTGGGACACTT-3′ and bases 460–470; RNA analysis
5′-AGTGCCAGGCGTTGAAAAGA-3′). The PCR conditions were For quantitative PCR, total RNA was extracted from ASMC and
the same in the second round as in the first round reaction. AS52 cells by using TRIzol® (Invitrogen, USA) according to the
The quantity ofgpt gene amplification was analyzed by ele-c manufacturer’s instructions. Total RNA was quantified and
trophoresis on 1.2% agarose gels stained with EtBr. The 0.5 kb reverse transcribed into cDNA using random hexamer primer
PCR product was cut out and purified for DNA sequence ana-ly and a GScript first-strand synthesis kit (GeneDirex, USA). To
sis using the EasyPure PCR/Gel Extraction Kit (Bioman Scientific assess the quantity ofc-Myc, Rab27b, Pak4, Ccnd1, Ccnd2, Arf1,
CO., LTD, Taiwan). DNA sequencing was carried out by ABI 3730 Src and Vegf in AS52 and ASMC cells, the primers used for
DNA Analyzer. the amplification were as follows in Supplementary Table S1.
qPCRBio syGreen Mix Lo-ROX reagent (PCR Biosystems, UK)
Cell invasion and soft agar colony forming assays was employed for fluorescence-based quantification of cDNA
The cell invasion was measured in the Corning Transwell® by real-time QPCR using a myGO PCR® (IT-IS Life Science Ltd,
24-well insert plates using Polyester membrane (μ8 m pore Ireland). The QPCR product was evaluated and normalized to
size). AS52-mutant cell (ASMC) were plated at×5 105 cells/well 18S ribosomal mRNA levels and the negative control.
in six-well plate and starved in the medium for additional 24 h
incubation. Starved cells were suspended in 0.1 ml serum-free Microarray
medium and added to the upper chamber coated with 0.1% For microarray experiment, Mouse miRNA OneArray® v6
basement membrane extract; 0.6 ml of regular medium was (Phalanx Biotech Group, Taiwan), which is based on Mouse
added to the lower chamber. After 24 h incubation, the cells (miRBase Release 20) and contain 1883 unique miRNA probes
subsided on the polyethylene terephthalate membrane in the and 114 experimental control probes. The detailed descriptions
bottom chamber were fixed with 4% paraformaldehyde (PFA) of the gene array list are available frohmttp://www.phalanx.
and stained with 0.05% crystal violet. The invasion ability of com.tw/products/MRmiOA_Probe.php . Fluorescent targets
these mutant cells were quantified by counting the cells thatwere prepared from 2 μg total RNA samples using miRNA
ULSTM Labeling Kit (Kreatech Diagnostics, The Netherlands). MEHP is genotoxic and mutagenic in AS52 cells
Labeled miRNA targets enriched by NanoSep 100K (Pall Because MEHP-induced formation of ROS led to DNA damage,
Corporation, USA) were hybridized to the Mouse miRNA we used the comet assay to evaluate whether MEHP causes
OneArray® with Phalanx hybridization buffer by using the DNA single-strand breaks. We used concentrations of 0, 10, 25
OneArray® Hybridization Chamber. The slides scanned using and 50 mM MEHP to carry out the experiment, though at 50 mM
a DNA Microarray Scanner (Molecular Devices, USA). The Cy5 MEHP all the cells died during collection. As shownFingure 2A,
fluorescent intensities of each spot were analyzed by GenePix MEHP indeed caused significant single-strand breaks in the
4.1 software (Molecular Devices). The signal intensity of each cells, and the addition of NAC could minimize the DNA damage
sample was analyzed by using R language. StatisticaPlvalue caused by MEHP, although the protection was not significant at
was calculated by student’ts-test and fold change was ou-t 25 mM MEHP. We then conducted a comet assay with the
PARPputted for evaluating differentially expressed genes. The -cri 1-KD AS52 cells and the XPD-KD AS52 cells, and as shown in
teria with fold change ≥ 1.74 or ≤ −1.74 anPd `value ≤ 0.05 are Figure 2B, even after exposure to a lower dose of MEHP (1 mM),
strongly recommended for further analysis. Sequentially, we the PARP-1-KD cells have a higher level of single-strand breaks.
used web-based tool mirsystem and KEGG pathway (database For the experiments involving mutation selection, AS52 that
for annotation, visualization and integrated discovery) to i-den we used was transformed from AA8 that have a defect in the
tify potential biological pathways for miRNAs showing similar HPRT gene by transferring theE.coli gpt gene into the cells with
expression profiles in cluster analysis. To develop the overall pSV2gpt. In this experiment, we exposed the cells to 0, 1, 10 and
network between miRNA and target genes, an enrichment 25 mM MEHP and added NAC to suppress the formation of ROS.
interaction analysis was performed using Cytoscape v3.1.0 We used colonies that have spontaneous and MEHP-induced
based on the gene ontology terms. mutations to produce mutations using 6-TG and then selected
the mutated colonies. The resultsFi(gure 2C) show that MEHP
mutagenicity rises as the dosage rises, while NAC lowers the
Results mutagenicity of MEHP by reducing the formation of ROS. This
MEHP-induced ROS cause genotoxicity in CHO cells indicates that MEHP indeed causes DNA damage and induces
cell mutation through the formation of ROS. We also carried out
As shown in Figure 1A, MEHP induces cells to generate intrac-el mutation selection on PARP-1-KD-AS52 and XPD-KD-AS52 and
lular ROS, and the amount of ROS generated is dependent on found that MEHP, even at 1 mM, has significant mutagenicity
the dose of MEHP. We next determined the source of the ROS. for PARP-1-KD-AS52, while the siRNA-control cells exposed to
Using the MitoSOX protocol, we found that after adding MEHP, 10 and 25 mM MEHP all died during the process of mutation
the mitochondria generate a large amount of superoxide anion selection. The above result shows that MEHP may induce cell
(17) (Figure 1B). Large amounts of ROS can cause oxidation of mutation through ROS. Furthermore, when PARP-1 function was
lipids, which increases the level of oxidative stress in cells. As blocked, the mutagenicity of MEHP toward the cells increased.
shown in Figure 1C, the amounts of malondialdehyde formation We subsequently did some gpt gene sequencing to
anaand of ROS formation caused by MEHP are proportional, in that lyze the mutation points on the genes. We randomly picked 25
they both increase as the dose of MEHP increases. If NAC was mutant colonies from the experimental group and named the
added to the experimental group, both the oxidative stress and cell strains ASMCs (AS52 mutant cellsF)i.gure 2D and E show
lipid peroxidation induced by MEHP were reduced. that the most common mutations were single base pair subs-ti
To evaluate whether the ROS induced by MEHP causes cy-to tutions, which comprised 90% of all mutations. Of all the tra-nsi
toxicity and genotoxicity to nucleotides, we used AA8 as the tion mutations, G:C to A:T mutations were the most common,
normal gene repair group and UV5 [nucleotide excision repair- comprising 22% of all such mutations.
deficient CHO cells] 1(4) and EM9 [base excision repair
(BER)deficient CHO cells]1(8) as the groups that are defective with
regard to repair of damaged genes. The results for cytotoxicity The mutant AS52 cells have carcinogenic properties
show that after cells were exposed to DEHP for 2 h, AA8 was not We used in vivo and in vitro experiments to test if ASMCs are
affected after DEHP was removed from the cell culture medium. carcinogenic. Using a soft agar assay to measure the anch-or
UV5 and EM9, however, showed a decrease in cell viability of age-independent growth ability of ASMCs is one characteristic
~25% at 50 mM DEHP, while cell viability of AS52 was decreased of transformation of a cell into a cancer ce2ll1).( The results
by ~40% (Figure 1D). Twenty four hours after DEHP was removed, show that the number of colonies formed by ASMCs was si-g
all four cell lines still retained 80% viability. In contrast, aftenrificantly greater than for AS52. Morphologically, we observed
the cells were exposed to MEHP for 2 h, there was a significant that the colonies formed by ASMCs were scattered over a wider
decrease in viability for all 4 cell lineFsig(ure 1E), particularly area (Figure 3A and B). The result of a cell invasion assay shows
for base excision repair-deficient EM9 cells after treatment with that ASMCs had significantly better invasion ability than AS52
10 mM MEHP. If NAC was added along with MEHP and then both (Figure 3C and D). In order to determine if ASMCs are onc-o
were removed after 2 h, cell viability was higher for all four cellgenic, we randomly picked 20 strains of ASMCs to inject into the
lines (Figure 1F). hypodermis of the legs of nude mice. After 4 weeks of
To further verify whether the MEHP-induced ROS would tion, as shown inFigure 3E, out of 20 strains of cells, 4 strains of
have the same genotoxicity effects in AS52 cells that have been ASMCs developed into tumors. The hematoxylin and eosin stain
altered to be nucleotide excision repair-knockdown (KD) and (Figure 3H) shows that the cells in the tissues had an irre-gu
BER-KD, we used siRNA to suppress the XPD (19) and PARP-1 lar appearance and alterations in their karyoplasmic ratio. In
(20) functions of the AS52 cellsF.igure 1G shows that at 10 mM addition, when immunohistochemistry was used to analyze
MEHP there is higher cytotoxicity in the PARP-KD AS52 cells. As the proportion of tumor protein c-myc entering the cell nucleus
a result, the metabolite MEHP does indeed have a higher cy-to (22,23), the results showF(igure 4Hand I) that inside the tumors,
toxicity than DEHP, and the ROS formed by MEHP have a higher ~60% of the c-myc moved into the nucleus, whereas 80% of the
toxicity toward cells that lack the BER system. c-myc moved into the nucleus in the positive control. The above
results show that MEHP indeed has carcinogenic potential in signaling pathway and Wnt signaling pathway), which may play
mammalian cells. a role in cancer induction (pathways in cancer, prostate cancer,
melanoma, chronic myeloid leukemia, pancreatic cancer, small
microRNA array and gene ontology enrichment cell lung cancer, endometrial cancer and renal cell carcinoma).
analysis of AS52 mutant cells Last, we focused on the ErbB signaling pathway, because this
We carried out microRNA array analysis with ASMCs to eva-lu pathway is most closely correlated with ovarian canc2e5r)(. As
ate their microRNA profile. As shown inFigure 4A, ASMCs had described by the results in a number of publicationsTa( ble 1),
106 microRNAs that were affected compared with normal AS52 we sequentially analyzed the interaction between the mi-cro
cells. Of these, the 45 downregulated microRNAs that were RNAs and the target genesF(igure 5A). Previous research has
selected for the heat map result, with a fold change < −0.8 and shown that various microRNAs are able to suppress the expre-s
P ≤ 0.0001 as our filter conditionsFi(gure 4B). The one showing sion of target genes E(RBB2, ERBB3, myc), but in ASMC cells,
the greatest downregulation was miR-130a-3p, which declined many microRNAs have downregulated expression, including
1.7 times compared to the control group. This has relevance for let-7 family and miR-125b-5p. This result suggests that the t-ar
the oncogenicity of mir-130a-3p in growth control, migration, get genes of mutated cells induced after exposure to the pla-sti
and invasion as mentioned in many publications 2(4). We then cizer metabolite MEHP are no longer suppressed by microRNA,
used the website miRSystem to carry out gene ontology enri-ch causing the cells to become cancerous. To assess the target
ment analysis. As shown in Figure 4C, the 45 downregulated mRNA expression of mutated cells, we performed quantitative
microRNAs are correlated with 5615 genes and regulate 22 pa-th PCR analysis in ASMC. We found almost 8-fold change ocf-myc
ways in the cellsP( ≤ 0.0001), including cell proliferation pa-th expression and 3-fold change ofRAS expression in the ASMC
ways (MAPK signaling pathway, ErbB signaling pathway, TGβF- (Figure 5B).
Discussion NAC (a ROS scavenger), we directly counteracted the ROS and
increased the formation of the antioxidant protein glutathione
Previous research indicated that DEHP, through its primary (31), which provided antioxidant capacity to the cells and sig n-ifi
(MEHP) or secondary (MEHHP and MEOHP) metabolites, induces cantly suppressed the induction of oxidative stress. Therefore,
the formation of ROS 2(,26) and causes alterations in the status we concluded that MEHP induces oxidative stress and causes
of cells through molecular signaling pathway2s7)(, especially DNA peroxide damage.
when genes are affected by environmental factors. This can lead Excess ROS cause oxidative stress and lead to DNA damage
to deleterious effects such as DNA adducts, DNA strand breaks, and peroxidation of proteins and lipids 3(2). Previous work has
DNA methylation or histone modifications in various tissues, shown that MEHP induces DNA oxidation and causes the f-or
resulting in gene mutations and changes in cell epigenetics that mation of 8-oxo-deoxyguanosine (6,33), base excision damage
can ultimately transform normal cells into cancer ce2ll8s–3(0). (34) and DNA single-strand breaks3(5). In our data shown that
Although the potential of DEHP in carcinogenesis has been well MEHP induced DNA single breaks; moreover, the defection of
investigated, there is little report clarifying the role of MEHP-base excision repair may increase the level of DNA single-strand
induced DNA damage in cell transformation from oxidative breaks in lower dose MEHP. Apparently, MEHP induce ROS pr-o
perspective. duction lead to DNA damage. In addition, MEHP is genotoxic to
In this study, AS52 CHO cells were used to test if MEHP cells that lack nucleotide excision repair, which may cause the
causes base excision damage and is mutagenic and then to eva-l formation of DNA adducts 3(6). This result also indicates that
uate the possibility of oncogenicity of the mutated cells and the MEHP not only causes DNA oxidation damage but also nucl-eo
involvement of microRNA epigenetics. Previous work used this tide damage.
method to study 3,5-dimethoxyamphetamine and the genotox- In the experiment analyzing mutagenicityF(igure 2),
shorticity of its metabolites1(4). In Figure 1, acute high-concentration term acute exposure to 25 mM MEHP increased the mutation
(10 mM) exposure to the metabolite MEHP gives rise to a lower rate about 3-fold, and the mutagenicity of MEHP is concentration
cell viability for CHO cells lacking base excision repair. By adding
dependent. Addition of NAC suppressed the formation of ROS conducted cell transformation analysis with a soft agar assay.
and lowered the mutation rate of MEHP. However, it is worth -not The experimental result shows that the number of colonies
ing that 1 mM MEHP in BER-deficient AS52 cells and 25 mM MEHP formed by the mutated cells (ASMCs) was greater than for AS52
in regular AS52 cells caused similar cell mutation rates, and the cells and with more scattered morphology, and the area in
comet assay result also shows that when cells lack BER, a low which the mutated cells were distributed was larger than for
concentration of MEHP can cause significant DNA single-strand the AS52 cells F(igure 3). In the invasion assay, ASMCs showed
breaks at a level of ~20%. In the result gopft gene sequencing greater capacity for motility and invasion. The reason for this
(Figure 2D), the mutations which MEHP caused the most were may be that by alteringc-myc, MEHP can cause expression and
base substitutions, with G:C base pairing being the highest activation of TIMP2 and upregulation of MMP2, which increases
in number. MEHP also produced ~22% G:C to A:T transitions. the cells’ capabilities for invasion38(,39). For thein vivo assay,
Generally speaking, the mutation of G:C base pairing is asso-ci as shown in Figure 3E, we used a xenograft model to determine
ated with oxidation damage3(7,38). Our results found that occ-ur the oncogenic ability of mutated cells. We randomly chose 20
rence of oxidative DNA damage could lead to base excision DNA strains of ASMCs, implanted them through subcutaneous inje-c
damage and mutagenicity, while the NAC treatment showed tion, measured the first tumor on day 14, and made observations
the opposite effects, strongly suggesting that DNA base excision until day 28. We found that 4 out of 20 nude mice had tumor
damage might be increased the frequency of cell mutation. angiogenesis. Hence, the oncogenic rate of ASMCs was 20%.
We transferred the mutated cells into culture dishes in-reg Epigenetic gene ataxia is an important process in ca-n
ular medium with 6-TG, cultured them for 14 days and then cer cells, which includes differentiation, proliferation, and
apoptosis. MicroRNAs may play a role as tumor suppressors in of miR-125b-5p and increase the expression of ERBB2 and
a cell and may be involved in the angiogenesis and metastasis ERBB3 through DNA methyltransferase 149(). Ovarian cancer
of a tumor. In recent years, it was found that if normal cells research also shows that in 19 out of 20 patients, miR-125b is
lack certain features of microRNA, this may lead to the form-a significantly decreased5(0). Herein, epigenetic modification of
tion of cancer cells4(0), and the reasons for this may include microRNAs enhanced activation of oncogene,c-myc and Ras,
mutation, epigenetic silencing, and alterations in microRNA and some pathway known to be ERBB targets such asPak4, Src
processing (41). As shown in Figure 4, unlike AS52, ASMCs have and Vegf (51,52) and c-Myc targets such asCcnd1, Ccnd2 and Arf1
45 microRNAs with significant downregulationP (≤ 0.0001), (43) (Figure 5B). Overall, consistent with previous ovarian cancer
and these include let-7 family members. This result is sim-i studies, we found that epigenetic modification of microRNA in
lar to the let-7 family, which plays a role in tumor suppr-es MEHP-induced mutated cells may have the same properties in
sion, as previously reported, and is able to suppress oncogenes ovarian cancer.
including the expression ofRAS (42) and c-myc (43). Therefore, Previous publications have shown that endocrine disruptor
expression of let-7 family is low in typical cancer cell4s4(,45). phthalates have the ability to change epigenetic expression. Our
Moreover, a study of ovarian cancer shows that 31 out of 47 research is the first to show that through the induction of- oxi
ovarian cancer patients have excessive expression ocf-myc, at dative stress in cells, the primary metabolite MEHP causes base
65.9% (46). Plisiecka-Halasaet al. also pointed out that 70% of excision damage to CHO DNA, which leads to mutagenicity. The
the cancer cells in ovarian cancer have excessive expression resulting mutations have oncogenic ability, which may be due to
of c-myc. changes in microRNA expression that cause ataxia in oncogenes
In the gene enrichment analysis F(igure 5), we confirmed and in tumor suppressor genes and lead to the formation of ca-n
that many microRNA in ASMCs are directly related to cancer- cer cells. In the future, we will further examine whether MEHP
associated pathways, including downregulated p53 and upr-eg affects the accuracy of DNA repair through regulation of ep-ige
ulated ErbB, Wnt and MAPK. Of the five microRNAs that show netic expression, and whether increased phthalate enhances the
the greatest difference from control cells (miR-130a-3p, miR- risk of cancer for humans by changing microRNA expression.
27a-3p, miR-25-3p, miR-92a-3p and miR-125b-5p), we found that
both miR-125b-5p and miR-27a-3p are related to the ErbB sig-n
aling pathway 4(7). Research also shows that the ERBB rece-p Funding
tor of ovarian cancer has excessive expression4(8). ROS may Ministry of Science and Technology in Taiwan
(NSC-102-2320-Bcause DNA methylation and thereby decrease the expression 033-001-MY3 and NSC-102-2320-B-033-004).
Acknowledgements 26. Erkekoglu, P.et al. (2010) Evaluation of cytotoxicity and oxidative DNA
damaging effects of di(2-ethylhexyl)-phthalate (DEHP) and
mono(2YJC and CYT designed the experiments. YJC and PYL carried ethylhexyl)-phthalate (MEHP) on MA-10 Leydig cells and protection by
out most of the experiments and analysis. MWC was the pri-n selenium. Toxicol. Appl. Pharmacol., 248, 52–62.
cipal investigator who conceived the study and responsible for 27. Kovacic, P. (2010) How dangerous are phthalate plasticizers? In-te
the manuscript writing. All the authors read and corrected the grated approach to toxicity based on metabolism, electron transfer,
manuscript. reactive oxygen species and cell signaling. Med. Hypotheses, 74, 626–
Conflict of Interest Statement: None declared. 628.
28. Singh, S. et al. (2012) Epigenetic effects of environmental chemicals bi-s
References phenol A and phthalates. Int. J. Mol. Sci., 13, 10143–10153.
29. Wu, S. et al. (2010) Dynamic effect of di-2-(ethylhexyl) phthalate on t-es
1. ATSDR (2002) Toxicological profile for di(2-ethylhexyl) phthalate. ticular toxicity: epigenetic changes and their impact on gene expr-es
2. Wang, W. et al. (2012) Mono-(2-ethylhexyl) phthalate induces oxid-a sion. Int. J. Toxicol., 29, 193–200.
tive stress and inhibits growth of mouse ovarian antral follicles. Biol. 30. Hsieh,T.H. et al. (2012) Phthalates stimulate the epithelial to mesenc-hy
Reprod., 87, 152. mal transition through an HDAC6-dependent mechanism in human
3. Tetz, L.M. et al. (2013) Mono-2-ethylhexyl phthalate induces oxidative breast epithelial stem cells. Toxicol. Sci., 128, 365–376.
stress responses in human placental cells in vitro. Toxicol. Appl. Ph-ar 31. Sun, S.Y. (2010) N-acetylcysteine, reactive oxygen species and beyond.
macol., 268, 47–54. Cancer Biol. Ther., 9, 109–110.
4. Takashima, K. et al. (2008) Different mechanisms of DEHP-induced 32. Valko, M.et al. (2007) Free radicals and antioxidants in normal phy-si
hepatocellular adenoma tumorigenesis in wild-type and Ppar alpha- ological functions and human disease. Int. J. Biochem. Cell Biol., 39,
null mice. J. Occup. Health, 50, 169–180. 44–84.
5. Ito, Y. et al. (2008) PPARalpha- and DEHP-induced cancers. PPAR Res., 33. Yang, G.et al. (2015) Mono(2-ethylhexyl) phthalate induces apoptosis in
2008, 759716. p53-silenced L02 cells via activation of both mitochondrial and death
6. Yang, G.et al. (2012) MEHP-induced oxidative DNA damage and apopto- receptor pathways. Environ. Toxicol., 30, 1178–1191.
sis in HepG2 cells correlates with p53-mediated mitochondria-depen-d 34. Çağlayan, M.et al. (2015) Oxidant and environmental toxicant-induced
ent signaling pathway. Food Chem. Toxicol., 50, 2424–2431. effects compromise DNA ligation during base excision DNA repair.
7. Rusyn, I. et al. (2001) Phthalates rapidly increase production of reactive DNA Repair (Amst.), 35, 85–89.
oxygen species in vivo: role of Kupffer cells. Mol. Pharmacol., 59, 744–750. 35. Kleinsasser, N.H. et al. (2004) Mono(2-ethylhexyl)phthalate exhibits
8. Bauer, N.C.et al. (2015) The current state of eukaryotic DNA base da-m genotoxic effects in human lymphocytes and mucosal cells of the
age and repair. Nucleic Acids Res., 43, 10083–10101. upper aerodigestive tract in the comet assay. Toxicol. Lett., 148, 83–90.
9. Gillet, L.C.et al. (2006) Molecular mechanisms of mammalian global 36. Iyama, T. et al. (2013) DNA repair mechanisms in dividing and
nongenome nucleotide excision repair. Chem. Rev., 106, 253–276. dividing cells. DNA Repair (Amst.), 12, 620–636.
10. Johnson, G.E. (2012) Mammalian cell HPRT gene mutation assay: test 37. Neeley, W.L. et al. (2006) Mechanisms of formation, genotoxicity, and
methods. Methods Mol. Biol., 817, 55–67. mutation of guanine oxidation products. Chem. Res. Toxicol., 19,
11. Stout, J.T.et al. (1985) HPRT: gene structure, expression, and mutation. 491–505.
Annu. Rev. Genet., 19, 127–148. 38. Yao, P.L. et al. (2012) Mono-(2-ethylhexyl) phthalate (MEHP) promotes
12. Phillips, B.J.et al. (1990) Chinese hamster ovary (CHO/HPRT) cell mut-a invasion and migration of human testicular embryonal carcinoma
tion assays with EMS, benzo[a]pyrene and benzidine. Mutagenesis, 5 cells. Biol. Reprod., 86, 160, 1–160,10.
(suppl), 15–20. 39. Yao, P.L. et al. (2011) Transcriptional suppression of Sertoli cell Timp2
13. Stankowski, L.F. Jret al. (1986) Quantitative and molecular analyses of in rodents following mono-(2-ethylhexyl) phthalate exposure is re-gu
ethyl methanesulfonate- and ICR 191-induced mutation in AS52 cells. lated by CEBPA and MYC. Biol. Reprod., 85, 1203–1215.
Mutat. Res., 160, 133–147. 40. Garzon, R. et al. (2009) MicroRNAs in cancer. Annu. Rev. Med., 60,
14. Chao, M.W. et al. (2012) Genotoxicity of 2,6- and 3,5-dimethylaniline in 167–179.
cultured mammalian cells: the role of reactive oxygen species. Toxicol. 41. Garzon, R. et al. (2006) MicroRNA expression and function in cancer.
Sci., 130, 48–59. Trends Mol. Med., 12, 580–587.
15. Wattanawaraporn, Re.t al. (2012) AFB(1) -induced mutagenesis of the 42. Johnson, S.M. et al. (2005) RAS is regulated by the let-7 microRNA fa-m
gpt gene in AS52 cells. Environ. Mol. Mutagen., 53, 567–573. ily. Cell, 120, 635–647.
16. Kim, H.W. et al. (2003) Mutagenicity of reactive oxygen and nitrogen 43. Psathas, J.N. et al. (2014) MYC and the rrt of MicroRNA maintenance.
species as detected by co-culture of activated inflammatory leukocytes Cold Spring Harbor Perspect. Med., 4, a014175.
and AS52 cells. Carcinogenesis, 24, 235–241. 44. Yanaihara, N.et al. (2006) Unique microRNA molecular profiles in lung
17. Cadenas, E.et al. (2000) Mitochondrial free radical generation, oxidative cancer diagnosis and prognosis. Cancer Cell, 9, 189–198.
stress, and aging. Free Radic. Biol. Med., 29, 222–230. 45. Iorio, M.V. et al. (2005) MicroRNA gene expression deregulation in
18. Stearns, D.M.et al. (2005) Uranyl acetate induces hprt mutations and human breast cancer. Cancer Res., 65, 7065–7070.
uranium-DNA adducts in Chinese hamster ovary EM9 cells. Mutage-n 46. Chen, C.H. et al. (2005) Overexpression of cyclin D1 and c-Myc gene
esis, 20, 417–423. products in human primary epithelial ovarian cancer. Int J Gynecol
19. Constantinescu-Aruxandei, De.t al. (2016) Mechanism of DNA loading Cancer, 15, 878–883.
by the DNA repair helicase XPD. Nucleic Acids Res., 44, 2806–2815. 47. Wu, X. et al. (2013) Coordinated targeting of the EGFR signaling axis by
20. De Vos, M. et al. (2012) The diverse roles and clinical relevance of PARPs microRNA-27a*. Oncotarget, 4, 1388–1398.
in DNA damage repair: current state of the art. Biochem. Pharmacol., 48. Lafky, J.M. et al. (2008) Clinical implications of the ErbB/epidermal
84, 137–146. growth factor (EGF) receptor family and its ligands in ovarian cancer.
21. Borowicz, S. et al. (2014) The soft agar colony formation assay. J. Vis. Biochim. Biophys. Acta, 1785, 232–265.
Exp., 92, e51998. 49. He, J. et al. (2012) Reactive oxygen species regulate ERBB2 and ERBB3
22. Dang, C.V. (1999) c-Myc target genes involved in cell growth, apoptosis, expression via miR-199a/125b and DNA methylation. EMBO Rep., 13,
and metabolism. Mol. Cell. Biol., 19, 1–11. 1116–1122.
23. Wade, M. et al. (2006) c-Myc, genome instability, and tumorigenesis: the 50. Nam, E.J. et al. (2008) MicroRNA expression profiles in serous ovarian
devil is in the details. Curr. Top. Microbiol. Immunol., 302, 169–203. carcinoma. Clin. Cancer Res., 14, 2690–2695.
24. Liu, Y. et al. (2016) MiR-130a-3p regulates cell migration and invasion 51. Arias-Romero, L.E. et al. (2010) A Rac-Pak signaling pathway is essential
via inhibition of Smad4 in gemcitabine resistant hepatoma cells. J. Exp. for ErbB2-mediated transformation of human breast epithelial cancer
Clin. Cancer Res., 35, 19. cells. Oncogene, 29, 5839–5849.
25. Maihle, N.J. et al. (2002) EGF/ErbB receptor family in ovarian cancer. 52. Holbro, T.et al. (2003) The ErbB receptors and their role in cancer p-ro
Cancer Treat. Res., 107, 247–258. gression. Exp. Cell Res., 284, 99–110.