Metabolic signatures of bisphenol A and genistein in Atlantic salmon liver cells

Accepted Manuscript Metabolic signatures of bisphenol A and genistein in Atlantic salmon liver cells Pål A. Olsvik, Kaja H. Skjærven, Liv Søfteland PII:

S0045-6535(17)31489-3

DOI:

10.1016/j.chemosphere.2017.09.076

Reference:

CHEM 19949

To appear in:

ECSN

Received Date: 13 July 2017 Revised Date:

14 September 2017

Accepted Date: 16 September 2017

Please cite this article as: Olsvik, På.A., Skjærven, K.H., Søfteland, L., Metabolic signatures of bisphenol A and genistein in Atlantic salmon liver cells, Chemosphere (2017), doi: 10.1016/ j.chemosphere.2017.09.076. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

RI PT

ACCEPTED MANUSCRIPT

Metabolic signatures of bisphenol A and genistein in Atlantic

M AN U

SC

salmon liver cells

AC C

EP

TE D

Pål A. Olsvik1,2*, Kaja H. Skjærven1, Liv Søfteland1

*Corresponding author

Tel: +47 41459367. Fax: +47 55905299 E-mail: [email protected]

1

National Institute of Nutrition and Seafood Research (NIFES), Bergen, Norway

2

Faculty of Biosciences and Aquaculture, Nord University, Bodø, Norway (current address)

1

ACCEPTED MANUSCRIPT

Abstract

Screening has revealed that aquafeeds with high inclusion of plant material may contain small amounts of endocrine disrupting agricultural pesticides. In this work, bisphenol A (BPA) and

RI PT

genistein (GEN) were selected as model endocrine disrupting toxicants with impact on DNA methylation in fish. Atlantic salmon hepatocytes were exposed in vitro to four concentrations of BPA and GEN (0.1, 1.0, 10 and 100 µM) for 48 h. Toxicity endpoints included cytotoxicity, global DNA

SC

methylation, targeted transcriptomics and metabolomic screening (100 µM). GEN was not cytotoxic in concentrations up to 100 µM, whereas one out of two cell viability assays indicated a cytotoxic

M AN U

response to 100 µM BPA. Compared to the control, significant global DNA hypomethylation was observed at 1.0 µM BPA. Both compounds upregulated cyp1a1 transcription at 100 µM, while estrogenic markers esr1 and vtg1 responded strongest at 10 µM. Dnmt3aa transcription was downregulated by both compounds at 100 µM. Metabolomic screening showed that BPA and GEN resulted in significant changes in numerous biochemical pathways consistent with alterations in

TE D

carbohydrate metabolism, indicating perturbation in glucose homeostasis and energy generation, and glutamate metabolism. Pathway analysis showed that while the superpathway of methionine degradation was among the most strongly affected pathways by BPA, GEN induced changes to

EP

uridine and pyrimidine biosynthesis. In conclusion, this mechanistic study proposes metabolites associated with glucose and glutamate metabolism, glucuronidation detoxification, as well as

AC C

transcriptional markers cyp1a1, vtg1, esr1, ar, dnmt3aa, cdkn1b and insig1 as markers for BPA and GEN exposure in fish liver cells.

Keywords: Atlantic salmon hepatocytes; Bisphenol A; Genistein; Metabolomics; Gene expression. Global DNA methylation

2

ACCEPTED MANUSCRIPT

1. Introduction

Atlantic salmon (Salmo salar) fillet, with its high content of fat, tend to accumulate lipid-soluble contaminants. Concerns have therefore been raised regarding human health risks of farmed salmon

RI PT

consumption (Hites et al., 2004; Nøstbakken et al., 2015). The diet is the main source of organic contaminants for farmed Atlantic salmon. Historically, these compounds were introduced mainly with marine ingredients such as fish oils (Berntssen et al., 2010). Today, with fishmeal and fish oil

SC

being increasingly replaced with plant protein and vegetable oils, contaminants might also stem from other sources. Potential sources include chemicals used in agriculture, as well as contaminants

M AN U

originating from food processing techniques, transport and storage. Farmed salmon might also be exposed to non-feed based contaminants during the production cycle. For example, modern Atlantic salmon aquaculture employ a relatively large amount of plastics in rearing tanks and piping, which has the potential to introduce plasticizers such as bisphenol A (BPA) into system water.

TE D

Recent screening has shown that novel Atlantic salmon aquaculture feeds, which consist of about 70% plant ingredients, may contain trace amounts of pesticides (Berntssen et al., 2010; NacherMestre et al., 2014). In an effort to study the potential toxicological impact of contaminants

EP

associated with present-day salmon farming, we have used in vitro models to search for biomarkers of exposure in fish. This research has focused on compounds such as endosulfan, chlorpyrifos and

AC C

pirimiphos-methyl (Krøvel et al., 2010; Søfteland et al., 2014; 2016; Olsvik et al., 2015a; 2017). Some of these chemicals, especially those that act as endocrine disruptors, may potentially impact mechanisms linked to DNA methylation. Feeds mainly based on plant ingredients may contain lower levels of B-vitamins and some indispensable amino acids compared to traditional feeds based on marine ingredient (Hansen et al., 2015; Hemre et al., 2016). Diets with suboptimal concentrations of nutrients necessary for folate-centered one carbon (1C) metabolism may thus render the farmed salmon more vulnerable to chemicals known to impact DNA methylation mechanisms (Dolinoy et al., 2007).

3

ACCEPTED MANUSCRIPT

Two of the best-studied toxicants known to affect DNA methylation are BPA and genistein (GEN) (Dolinoy et al., 2006; Dolinoy et al., 2007). Both chemicals are considered to be weak endocrine disruptors (Krishnan et al., 1993; Patisaul and Adewale, 2009). BPA is a ubiquitous environmental

RI PT

contaminant originating mainly from polycarbonate plastics and epoxy resins (Staples et al., 1998). As an endocrine disruptor, BPA has been shown to bind to estrogen receptors (ERs), resulting in feminizing effects in fish and other animals (Dolinoy et al., 2007). GEN is a plant-derived phytoestrogen with ability to bind to ER beta receptor, activate PPARs and Nrf2 pathway, and inhibit

SC

tyrosine kinases (Morito et al., 2001; Dolinoy et al., 2006; Fan et al., 2006; Kim et al., 2009). Hence, both chemicals may interfere with vertebrate reproduction. In cells, these compounds can bind to

M AN U

nuclear receptors and initiate transcription of a number of estrogen-responsive genes (Kennedy et al., 2014). In male fish, increased transcription of ER genes and vitellogenin (VTG) are typical biomarkers of such chemicals (Sumpter and Jobling, 1995). Cross-talk between ER and the aryl hydrocarbon receptor (AhR) may also result in altered transcription of cytochrome P450 genes

TE D

(Beischlag et al., 2008).

The aim of this study was to study the mechanistic effects of two endocrine disrupting toxicants

EP

affecting DNA methylation in Atlantic salmon liver cells. Based on their known mode of action, BPA and GEN were selected as model toxicants. Identified biomarkers will be applied as potential

AC C

markers in follow-up in vivo examinations of the impact of compounds associated with present-day salmon farming. Atlantic salmon hepatocytes were exposed to four concentrations of BPA and GEN (control, 0.1, 1.0, 10.0, 100 µM) for 48 h. Cytotoxicity was examined with the MTT assay and the xCELLigence system. Global DNA methylation was determined with a HPLC-based method. Molecular endpoints included targeted transcription and cellular metabolites. Eighteen potential markers for cellular stress and DNA methylation were selected for transcriptional analysis using RTqPCR. Cells exposed to the highest BPA and GEN concentration (100 µM), were selected for metabolomic profiling with GC/MS and LC/MS/MS platforms (polar and non-polar metabolites). Pathway analysis was used to search for causal relationships and biomarkers.

4

ACCEPTED MANUSCRIPT

2. Materials and methods

RI PT

2.1. Cell harvesting Atlantic salmon (Salmo salar) was maintained at the Industrilaboratoriet (ILAB) animal holding facility, Bergen, Norway. The fish were kept in flowing sea water at 7-9.5°C, 34.4‰ and a 12/12 light/dark cycle. Wastewater O2 levels were always above 7 mg/l and the pH was 8.1-8.2. The fish

SC

were fed once a day with a special-made feed produced without synthetic antioxidants and with low levels of contaminants, delivered by EWOS, Norway (Spirit 400-50A HH, 6.0 mm). Hepatocytes

M AN U

were isolated from male juvenile Atlantic salmon (n=5, mean±SEM: 214±7 g) with a two-step perfusion method described by Søfteland et al. (2009). The fish were sacrificed by terminal anaesthetization with tricaine methanesulfonate (MS-222) (200 mg/l). Fish sacrifice and harvesting of cells were conducted by the authors and approved by the Norwegian Animal Research Authority

TE D

(NARA) via NIFES' Animal Care and Use Committee.

Prior to exposure, the Trypan Blue exclusion method was used to determine cells viability (Lonzo,

EP

Medprobe, Oslo, Norway). Cell suspensions were put on 5 µg/cm2 laminin (Sigma-Aldrich, Oslo, Norway) coated culture plates (TPP, Trasadingen, Switzerland) and the hepatocytes were kept at

AC C

10°C in a sterile incubator without additional O2/CO2 supply (Sanyo, CFC FREE, Etten Leur, Netherland). The following cell densities were used: A) for cytotoxicity evaluation in xCELLigence 96-well plates and regular 96-well plates for the MTT assay: 2×105 cells per well, and B) for RTqPCR and metabolite profiling: 7.2×106 cells per well in 6-well plates (in 3 mL complete L-15 medium).

2.2. BPA and GEN exposure Hepatocyte cells were cultured for 36-40 h prior to chemical exposure with exchange of medium after 18-20 h. Cells were kept as controls or treated with BPA and GEN (0.1, 1.0, 10 and 100 µM)

5

ACCEPTED MANUSCRIPT and harvested after 48 h exposure. Cells were also exposed to the demethylating agent 5-aza-2′deoxycytidine (0, 0.001, 0.01, 0.1, 1 and 10 µM 5-AZA), which was included in the study as a positive control. BPA, GEN and 5-AZA were dissolved in DMSO, with an equal amount of DMSO (0.1%) used in the control group. BPA, GEN and 5-AZA were obtained from Sigma (Sigma-Aldrich,

RI PT

Oslo, Norway). Hepatocyte cells were exposed in triplicate wells for RT-qPCR and metabolite profiling, and in 96-wells culture plates for the xCELLigence and MTT cytotoxicity screening (single wells). The exposure medium was exchanged after 18-20 h. The number of biological replicates was 5 for all analytical methods. The final cell pellet was resuspended in L-15 medium (Sigma Aldrich,

SC

Oslo, Norway) containing 10% FBS (Sigma Aldrich, Oslo, Norway), 1% glutamax (Invitrogen, Norway) and 1% penicillin–streptomycin–amphotericin (10,000 units/ml potassium penicillin,

M AN U

10,000 µg/ml streptomycin in sulfate and 25 µg/ml amphotericin B) (Lonzo, Medprobe, Oslo, Norway).

2.3. Cytotoxicity screening

TE D

Two methods were applied to determine the cytotoxicity of BPA and GEN (5-AZA was not evaluated for cytotoxicity). Cell viability was determined with the MTT method using the In Vitro Toxicology assay kit according to the manufacturer’s protocol (Sigma Aldrich) and by the

EP

xCELLigence system (Real-Time Cell Analyzer RTCA-SP, ACEA Biosciences, San Diego, USA) (Abassi et al., 2009). For the latter method, recording of cell index (CI) values and normalization was

AC C

performed using the RTCA Software version 1.2.1. The real-time cell monitoring was conducted at 10°C in an incubator without additional O2/CO2 supply (Sanyo, CFC FREE, Etten Leur, Netherland). Data was collected with 2 min intervals after BPA and GEN exposure for 12 h and then every 15 min for 60 h. For calculation of cell viability after 48 hours of exposure, the impedance signal was analyzed by normalizing data of each singe well to a reference time point set one hour before the final exchange of exposure medium. Determination of cytotoxic effect was done according to the International standardized test for in vitro cytotoxicity ISO 10993-5:2009 (ISO 2009).

2.4. Global DNA methylation

6

ACCEPTED MANUSCRIPT Cell culture pellets were defrosted in ATL lysis buffer and homogenized using a Precellys 24 homogenizer at 3 x 15 s at 6000 rpm with intervals of 10 sec. DNA extraction were performed according to Qiagen DNeasy Blood and Tissue kit (Qiagen, Crawley, UK). The quantity of DNA was measured using Qubit Fluorometric Quantifcation (Thermo Fisher Scientific, USA). DNA

RI PT

methylation level was determined using a modified HPLC method (Ramsahoye, 2002), as previously described by Skjaerven et al. (2014). The extracted DNA was digested to single nucleotides using DNA Degradase according to manufacturer’s instructions (Zymo Research). After enzymatic digestion, the samples were diluted to a volume of 60 µL in 1xTE buffer with the appropriate DNA

SC

concentration of 30 ng/µL, and stored at -20°C until HPLC detection. A dilution curve of known adenine, guanine, cytosine, thymine, methyl-cytosine and uracil nucleotide standard mix was

M AN U

analyzed prior to and after the DNA samples. Uracil was included in the standard mix as a reference for RNA free DNA. Chromeleon software (Thermo Fisher Scientific, USA) was used for data processing from the HPLC. Percent DNA methylation was calculated using molar equivalent for both cytosine (dCMP) and methyl-cytosine (5mdCMP). The molar equivalent is the peak area divided by

respectively).

EP

2.5. RNA isolation

TE D

the excitation coefficient (excitation coefficient: 9300 and 11800 for dCMP and 5mdCMP,

Total RNA from the hepatocytes was extracted with the RNeasy Plus mini kit (Qiagen, Crawley,

AC C

UK). RNA was eluted in 50 µl RNase-free MilliQ H2O and stored at -80°C before further processing. RNA quality and integrity was checked using the NanoDrop ND-1000 UV-Vis Spectrophotometer (NanoDrop Technologies, Wilmington, DE. USA) and the Agilent 2100 Bioanalyzer with the RNA 6000 Nano kit (Agilent Technologies, Palo Alto, CA, USA). RNA 260/280 and 260/230 nm ratios were 2.02 ± 0.00 and 2.24 ± 0.04 in cells, respectively (n=39, mean ± SEM). The Bioanalyzer RNA integrity number (RIN) was 10.0 ± 0.0 for seven randomly picked RNA samples used for RT-qPCR (mean ± SEM).

2.6. RT-qPCR

7

ACCEPTED MANUSCRIPT In total, 21 genes were quantified with RT-qPCR, including three reference genes. PCR assays are shown in Table 1. A two-step real-time RT-PCR protocol previously described by Olsvik et al. (2015b) was used to quantify the transcriptional levels of the selected genes. Briefly, total RNA input was 500 ng in each reaction. Quality controls “no template controls” (ntc) and “no amplification

RI PT

controls” (nac) were run for quality assessment for each PCR assay. Mean normalized expression (MNE) of the target genes was determined using a normalization factor based upon actb, eef1a1 and

AC C

EP

TE D

M AN U

SC

uba52 (M<0.17), as calculated by the geNorm software (Vandesompele et al., 2002).

8

ACCEPTED MANUSCRIPT

RI PT

Table 1. PCR assays.

Gene Potential marker for

Accession no.

Detoxification

AF210727

subfamily A member 1 Cytochrome P450 family 3

vtg1

subfamily A

Vitellogenin 1

Detoxification

DQ361036

Endocrine disruption

TE D

cyp3a

M AN U

Cytochrome P450 family 1 cyp1a1

Forward primer

SC

Gene name symbol

NM_001044897

ar

Estrogen receptor 2

Androgen receptor

EP

Endocrine disruption

c_salmon

AC C

esr2

Estrogen receptor 1

Endocrine disruption

Endocrine disruption

NM_180966

NM_001083123

efficiencyBPA/G

Reverse primer (bp)

EN

GGCATCCCGGTGAACTTTA

GGTGTTGGTTTTCGGTTTGG

114

2.06/2.03

146

1.92/2.03

209

1.93/1.83

112

1.86/2.03

123

2.01/2.07

117

1.82/2.11

107

2.06/1.90

A

ACTAGAGAGGGTCGCCAAG

TACTGAACCGCTCTGGTTT

A

G

GTCATCAATGAGGATCCAAA

GCCTCAGCGATCAGTGCA

GGCCA

CCAT

>Contig15966_Atlanti

esr1

PCR Amplicon size

TGGAGGTGATGCAGAGCT GGTCTCCCCAGCCAGTCATA

TCT GATTAACGGAGCGCCACA

TGATCAGCTGGGCCAAGAAG

TC

GGATGAGGTCGGAGCAGTT

GGCTCAATGGCCTCCAGA

C

AT

CCCAGATGGTCGTGAAAGGA TGAACCATGAGCCGGTCAT casp3b

Caspase 3B

Apoptosis

NM_131877

T

T

9

ACCEPTED MANUSCRIPT

Cell proliferation. differentiation fos

Fos proto-oncogene

and transcription regulation

CAAGTCCGGGCATGAAGA NM_205569

GGGTATTACCCGCTCAACCA

(hsp70)

Cellular stress

DY725986

Tumor necrosis factor tnfr

receptor

Cellular stress

NM_001141773

Membrane-associated progesterone receptor component 1

Glucose / Energy Metabolism

c_salmon

Regulation of cyclin-dependent

cdkn1b

Cyclin-dependent kinase

protein serine/threonine kinase

>Contig44819_Atlanti

Inhibitor 1B

activity

c_salmon (snudd)

Prostaglandinendoperoxide synthase 2

Lipid metabolism

AY848944

TE D

ptgs2

T

AAGACCTGCCTCCGTTGTAC

CTGAGGCACTCCCGTGTTT

A

C

TACATGGCCTGCAGCACTGT

M AN U

pgrmc1

>Contig22532_Atlanti

GTGCAGGCTGCCATCTTAGC

SC

hspa8

TGCCCCTCTGGGAATAACAA

ACAGCCCCCCGACTTACAAT

>Contig8978_Atlantic Insulin induced gene 1

Metabolism

_salmon (snudd)

EP

insig1

DNA (cytosine-5-)-

Maintenance DNA methylation methyltransferase 1 DNA (cytosine-5-)-

dnmt3aa methyltransferase 3A

synthase

108

1.92/2.03

140

2.08/

G

113

/1.82

151

1.82/1.98

112

1.94/1.82

111

1.93/2.07

152

2.01/2.01

112

1.82/2.04

81

1.92/2.09

AGCAGGCGGGTCTTTTTCT C GGTGTAGGGCAGTCCTTT GG AGCTGGGCATTATTGGCAA

GGAGCCCCACAAGTTCAAGA

A

CAAGTTCGGGGGTAGTGGT

TCTTGGCCTTGGACACCTT

C

C

TCACTGACCCCCATTGCAA

De novo DNA methylation

NM_001018134

GGCGCCTGTTCTTTGAGTTT

Transsulfuration

NM_001111232

CTTTGCCCTGGTGGTTCATG

Cystathionine-betacbs

2.06/1.92

GCGCTGCCGTACTGATACT

XM_014193379

AC C

dnmt1

102

CATGACCCCTCCAGCTGTC

RI PT

Heat shock protein A8

GA

ACCACTCCAAACACCATTT GC

10

ACCEPTED MANUSCRIPT

TGGTCTCCAGCTCCCATAT

Rate limiting enzyme, XM_014155171

adenosyltransferase 1A2

210

1.89/1.91

126

2.01/2.12

126

1.82/1.97

102

2.06/1.91

C

methionine cycle

Eukaryotic translation eef1a1

GCTGCTGTGTGGAGAGATCA

RI PT

Methionine mat1a2

CTCATGTCACGCACAGCAA

Refgen

AY422992

AGACAACCCCAAGGCTCTCA

elongation factor 1 alpha 1 Ubiquitin A-52 residue

A

TTGTTGGTGTGTCCGCACT

ribosomal protein fusion

Refgen

NM_001037113

AF057040

CGAGCAGGAGATGGGAACC

CAACGGAAACGCTCATTGC

TE D

Refgen

EP

Beta-actin

T

AC C

actb

M AN U

product 1

CGAGCCTTCTCTCCGTCAGT

SC

uba52

11

ACCEPTED MANUSCRIPT 2.7. Metabolomic screening Global biochemical profiles were determined in 15 Atlantic salmon hepatocyte cell culture samples, i.e. in 5 controls and 5 exposed to 100 µM BPA and 5 exposed to 100 µM GEN. Samples were extracted and prepared for analysis using Metabolon’s standard solvent extraction method. Metabolite

RI PT

profiling was conducted as previously described by Olsvik et al. (2017). Briefly, the extracted samples were split into equal parts for analysis on GC/MS and LC/MS/MS platforms. The LC/MS portion of the platform was based on a Waters ACQUITY UPLC and a Thermo-Finnigan LTQ mass spectrometer, which consisted of an electrospray ionization (ESI) source and linear ion-trap (LIT)

SC

mass analyzer. Samples destined for GC/MS analysis were analyzed on a Thermo-Finnigan Trace DSQ fast-scanning single-quadrupole mass spectrometer using electron impact ionization. Accurate

M AN U

mass determination and MS/MS fragmentation (LC/MS/MS) was based on Waters ACQUITY UPLC and a Thermo-Finnigan LTQ-FT mass spectrometer, which had a linear ion-trap (LIT) front end and a Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer backend. For ions with counts greater than 2 million, an accurate mass measurement could be performed. Accurate mass

TE D

measurements could be made on the parent ion as well as fragments. The typical mass error was less than 5 ppm. Instrument variability was 4% for internal standards and total process variability for endogenous metabolites was 12%. Identification of known chemical entities was based on comparison

EP

to metabolomic library entries of purified standards. The metabolomic screening was conducted by

AC C

employees at Metabolon. Inc. (Durham, NC, USA).

2.8. Statistics

The GraphPad Prism 6.0 software (GraphPad Software. Inc., San Diego. CA. USA) was used for statistical analyses of the cytotoxicity, global DNA methylation and gene expression data. One-way ANOVA with Holm-Sidak’s post-test analysis was used to compare exposed groups with the controls. Normality was checked with the Brown-Forsythe and Bartlett’s tests, and the data were logtransformed prior to ANOVA analysis when necessary. Outliers, as detected by ROUT method (ROUT

Q=1%), were omitted from the transcriptional dataset. The ROUT method is based on the False

12

ACCEPTED MANUSCRIPT Discovery Rate (FDR). With Q=1%, no more than 1% of the identified outliers are false. A significance level of P<0.05 was used for all tests.

To calculate differential metabolite expression, pathways enrichment analysis was done by

RI PT

annotation with the KEGG database (blastx E10-5). Following normalization to total protein (Bradford assay) and log transformation, ANOVA contrasts were used to identify biochemicals that differed significantly between experimental groups. Welch’s two-sample t-tests were used to identify biochemicals/metabolites that differed significantly between the experimental groups (P<0.05).

SC

Missing values from the metabolite screening were assumed to be below the limits of detection and these values were imputed with the compound minimum (minimum value imputation). Correction for

M AN U

multiple testing was done with FDR using q-values (P-adj) (Benjamini & Hochberg, 2005). Statistical analyses of the log-transformed data were performed with the program “R” (The Comprehensive R Archive Network). Functional pathway analyses were generated through the use of QIAGEN’s

Ingenuity®

Pathway

(IPA®,

QIAGEN

Redwood

City,

AC C

EP

TE D

www.qiagen.com/ingenuity).

Analysis

13

ACCEPTED MANUSCRIPT

3. Results

3.1. Cytotoxicity According to the MTT assay, BPA and GEN exposure for 48 h induced no cytotoxic effect on

RI PT

Atlantic salmon hepatocytes in concentrations up to 100 µM (Supplementary File 1, Fig. S1A and B). Based on impedance data and the xCELLigence system, BPA acted cytotoxic at 100 µM (one-way ANOVA, Holm-Sidak’s post hoc analysis, p=0.0203). Compared to the control, BPA exposure

3.2. Global DNA methylation

M AN U

determined with the xCELLigence system (Fig. S1D).

SC

reduced the cell index to 61% at 100 µM (Fig. S1C). No cytotoxic effect was seen for GEN as

Significantly reduced global DNA methylation level (hypomethylation) was observed in cells exposed to 1.0 µM BPA (one-way ANOVA, p=0.0181), but not at lower or higher concentrations (Supplementary File 1, Fig. S2A). The result signals a non-linear concentration-response pattern for

TE D

BPA. No significant effects of GEN (Fig. S2B) and 5-AZA (data not shown) exposure were seen on global DNA methylation levels in exposed cells.

EP

3.3. Transcriptional responses

Eight of 13 examined stress-responsive genes were differentially regulated in Atlantic salmon

AC C

hepatocytes exposed to BPA (Fig. 1). Cyp1a1 and cyp3a, markers for cytochrome P450 mediated detoxification, were significantly upregulated at 100 µM BPA (one-way ANOVA, Holm-Sidak’s post hoc analysis, Fig. 1A and B). BPA exposure affected markers for estrogenic effect at 1.0 µM (vtg1↑), 10 µM (vtg1↑, esr1↑) and 100 µM (vtg1↑, esr1↑, ar↓). Vtg1 showed the strongest effect with a non-linear concentration-response pattern, and was 24.2-fold upregulated in cells exposed to 10 µM BPA compared to the control (Fig. 1C). Esr1 was 3.6-fold and 3.4-fold upregulated in cells exposed to 10 and 100 µM BPA, respectively (Fig. 1D), while ar was downregulated at 100 µM BPA (Fig. 1F). Of the remaining genes, fos was significantly downregulated at the three lowest

14

ACCEPTED MANUSCRIPT concentrations but unaffected at 100 µM BPA (Fig. 1H), whereas insig1 was significantly upregulated at 10 µM BPA (Fig. 1M) and cdkn1b was upregulated at 100 µM BPA (Fig. 1K). The latter result, together with the vtg1 result, indicates a non-linear concentration-response pattern to

AC C

EP

TE D

M AN U

SC

RI PT

BPA exposure.

Figure 1. Concentration-response effect of bisphenol A on stress-responsive genes in Atlantic salmon hepatocytes.

15

ACCEPTED MANUSCRIPT Transcriptional responses of 17 selected genes in Atlantic salmon hepatocytes exposed to 0, 0.1, 1.0, 10 and 100 µM BPA. A) cyp1a1, B) cyp3a, C) vtg1, D) esr1, E) esr2a, F) ar, G) casp3b, H) fos, I) hspa8, J) tnfr, K) cdkn1b, L) ptgs2 and M) insig1. Bars with asterisks indicate statistical differences in mean values between control and BPA exposed hepatocytes (one-way ANOVA, Holm Sidak’s post-

RI PT

test. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001. Data are the mean±SEM, n=5.

BPA and GEN have both been classified as weak endocrine disruptors. The transcriptional concentration-response pattern of GEN exposure roughly resembled the response to BPA exposure in

SC

Atlantic salmon hepatocytes. Cyp1a1 was significantly upregulated by GEN at 100 µM (Fig. 2A), while cyp3a did not respond significantly to GEN exposure. Vtg1 and esr1 both showed the strongest

M AN U

response in cells exposed to 10 µM (Fig. 2C and D). Compared to the control, exposure to 0.1, 1.0, 10 and 100 µM GEN upregulated vtg1 2.6-fold, 11.8-fold, 17.2-fold and 5.0-fold, respectively. Esr1 showed a 2.4-fold, 3.4-fold and 1.9-fold upregulation at 1.0, 10 µM and 100 µM GEN, respectively. As opposed to BPA, GEN at 100 µM downregulated esr2a (Fig. 2E), but did not affect ar or fos

TE D

transcription significantly. GEN upregulated cdkn1b (Fig. 2K) and insig1 (Fig. 2M) in a similar

AC C

EP

fashion as BPA, except that the latter gene was significantly upregulated also by 100 µM.

16

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 2. Concentration-response effect of genistein on stress-responsive genes in Atlantic salmon hepatocytes.

Transcriptional responses of 17 selected genes in Atlantic salmon hepatocytes exposed to(0, 0.1, 1.0, 10 and 100 µM GEN. A) cyp1a1, B) cyp3a, C) vtg1, D) esr1, E) esr2a, F) ar, G) casp3b. H) fos, I) hspa8, J) pgrmc1, K) cdkn1b, L) ptgs2 and M) insig1. Bars with asterisks indicate statistical differences in mean values between control and BPA exposed hepatocytes (one-way ANOVA, Holm Sidak’s post-test. * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001. Data are the mean±SEM, n=5.

17

ACCEPTED MANUSCRIPT Four genes associated with DNA methylation and folate one-carbon metabolism were specifically examined (Fig. 3). Dnmt1, whose protein product has a role in maintenance methylation, was not significantly affected by BPA or GEN treatment at the studied concentrations. Both compounds downregulated dnmt3aa at 100 µM, while GEN also significantly downregulated dnmt3aa at 10 µM

RI PT

(Fig. 3C and D). In addition, GEN downregulated cbs at 100 µM (Fig. 3F), a response not seen in the

AC C

EP

TE D

M AN U

SC

cells after exposure to BPA.

Figure 3. Concentration-response effect of bisphenol A and genistein on DNA methylationrelevant genes in Atlantic salmon hepatocytes.

18

ACCEPTED MANUSCRIPT Expression of genes linked to DNA methylation in Atlantic salmon hepatocytes exposed to BPA (A.C.E.G) and genistein (B.D.F.H). A) and B) dnmt1, C) and D) dnmt3aa. E) and F) cbs and G) and H) mat1a2. Bars with asterisks indicate statistical differences in mean values between control and BPA exposed hepatocytes (one-way ANOVA, Holm Sidak’s post-test. * p<0.05, ** p<0.01, ****

RI PT

p<0.0001. Data are the mean±SEM, n=5.

Significant correlation between BPA exposure and transcription was found for estrogenic markers vtg1, esr1 and ar, as well as for DNA methylation markers dnmt3aa and cbs (Spearman’s rank

SC

correlation analysis). Significant correlation between GEN exposure and transcription was found for vtg1, esr1, esr2a, dnmt3aa and cbs. Exposure to the demethylating agent 5-AZA had no effect on the

3.4. Metabolomic screening

M AN U

transcription of the 17 evaluated genes (data not shown).

A total of 496 named biochemicals (compounds of known identity) were identified in liver cells in

TE D

the present study. Metabololic screening, which was conducted only for cells exposed to 100 µM, showed that BPA significantly upregulated 87 biochemicals and downregulated 33 biochemicals compared to the control, while GEN upregulated 19 and downregulated 46 biochemicals (Welch’s

EP

two-sample t-tests, p<0.05) (Fig. 4A). Table 2 shows a list of common biochemicals affected by both chemicals. Genes affected either by 10 or 100 µM of each chemical are included in Table 2.

AC C

Supplementary File 2 lists all significantly affected biochemicals.

Significantly elevated concentrations of BPA, BPA monosulfate and BPA glucuronide were found in Atlantic salmon hepatocytes exposed to BPA for 48 h compared to the control (Welch’s two-sample t-tests, p<0.05, using minimum value imputation for the controls) (Supplementary File 2). In terms of absolute values, BPA in the cells was present mainly in its original form or as its major metabolite (BPA:BPA glucuronide (major metabolite):BPA monosulfate; ratio 1.00:0.57:0.03). Significantly elevated concentration of GEN was found in cells exposed to GEN compared to the control (Welch’s two-sample t-tests, p<0.05, using minimum value imputation for the controls).

19

ACCEPTED MANUSCRIPT

UDP-glucuronate, the source of the glucuronosyl group in glucuronosyltransferase reactions, was the most strongly affected metabolite in cells exposed to BPA (Supplementary File 2). Compared to the control, exposure to 100 µM BPA reduced the levels of UDP-glucuronate 25-fold in Atlantic salmon

RI PT

hepatocytes. Phase II detoxification through glucuronosyltransferase reactions was also predicted by significantly reduced levels of UDP-galactose (-2.3-fold), UDP-glucose (-2.2-fold) and glucoronate (-4.6-fold), as well as uridine 5'-diphosphate (UDP) (-3.6-fold) and uridine (-1.5-fold), in cells exposed to BPA. A role for glucuronidation detoxification was also predicted for GEN in Atlantic

SC

salmon hepatocytes (Supplementary file 2). The level of UDP-glucuronate was 8.3-fold lower in GEN-exposed cells compared to the control. GEN exposure also significantly reduced the levels of

M AN U

UDP-galactose (-1.5-fold) and glucoronate (-1.6-fold), and uridine 5'-diphosphate (UDP) (-2.4-fold) and uridine (-1.3-fold) in Atlantic salmon hepatocytes.

Exposure to BPA resulted in perturbation in carbohydrate metabolism with significant effects on

TE D

glycogen catabolism (maltotetraose, maltotriose and maltose) and pentose phosphate pathway (6phosphogluconate, arabonate/xylonate) (Fig. 4B). GEN treatment showed overall similar outcomes on carbohydrate metabolism as BPA, with a significant decrease in glycolysis intermediates (e.g.,

EP

glucose, glucose 6-phosphate), pentose phosphate pathway and pentose sugars (6-phosphogluconate. arabonate/xylonate) and glycogenolysis products (maltotetraose, maltotriose and maltose) (Fig. 4B).

AC C

The treatment of hepatocytes with BPA and GEN thus resulted in significant changes in numerous biochemical pathways with consistent alterations in carbohydrate metabolism, indicating disruption in glucose homeostasis and energy generation. Both compounds affected glutamate metabolism (Fig. 4C). Treatment with BPA resulted in elevated levels of N-acetylglutamate and N-acetylglutamine. Exposure to GEN resulted in higher levels of glutamate and N-acetylglutamate.

Four glutathione metabolism biochemicals were increased by BPA, i.e. reduced glutathione (GSH), S-methylglutathione, 5-oxoproline and ophthalmate, a response not reproduced by GEN exposure. This finding indicates a differential ability to affect the anitoxidative defense (Supplementary File 2),

20

ACCEPTED MANUSCRIPT as also indicated by higher levels of vitamin E (alpha tocopherol) in BPA exposed cells. Elevated levels of GSH will enhance the cellular defense against oxidative stress, and may reflect a protective response. Several sphingolipid metabolism biochemicals were induced by BPA exposure but not by GEN exposure, indicating perturbation of membrane lipids or in mechanisms linked to sphingolipid

RI PT

signaling. Accumulation of fat (steatosis) was predicted as the main hepatotoxic outcome of both compounds. Finally, BPA and GEN appear to have different ability to impact vitamin B metabolisms. While BPA significantly induced thiamin (vitamin B1), thiamin monophosphate and

SC

pyridoxine (vitamin B6), GEN significantly induced pyridoxal phosphate.

Using a combination of significantly affected metabolites and genes as input, the top canonical

M AN U

pathways predicted induced by BPA and GEN exposure were different (IPA Core Analysis, using entities recognized by the Human Metabolome Database (HMDB), Table 3). Pathway analysis predicted that while the superpathway of methionine degradation was among the most strongly affected mechanism by BPA exposure, GEN treatment induced changes to uridine and pyrimidine

TE D

biosynthesis as well as to estrogen-mediated S-phase entry. Based on the predicted cascade of upstream transcriptional regulators that can explain the observed expression changes, IPA Core Analysis suggested that BPA and GEN acted on similar biological processes and molecular

EP

mechanisms. Three out of the four top predicted upstream regulators were common, i.e. EGFR, afatinib and sirolimus, indicating that the two compounds have relatively similar mode of action in

AC C

the cells (Table 3). Based on hepatotoxicity, similar mode of action was also predicted, with steatosis and morphological changes on the liver as the main effect.

Fig. 4A

21

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

Fig. 4B

Fig. 4C

22

RI PT

ACCEPTED MANUSCRIPT

SC

Figure 4. Effects of bisphenol A and genistein in Atlantic salmon hepatocytes.

A) Venn diagram of overlapping metabolites and genes in Atlantic salmon hepatocytes exposed to

M AN U

BPA and GEN for 48 h. The VENN input lists were generated by adding significant biochemicals in cells exposed to 100 µM BPA (120 biochemicals) and GEN (65 biochemicals) with genes showing significant response at 10 or 100 µM BPA (9 transcripts) and GEN (8 transcripts). B) Carbohydrate metabolism biochemicals affected by BPA and GEN exposure in Atlantic salmon hepatocytes. C)

TE D

Glutamate metabolism biochemicals affected by BPA and GEN exposure in Atlantic salmon hepatocytes. Arrows in the figure indicate the impact of BPA and GEN. In the tables, negative

AC C

EP

reciprocals are given for any value smaller than 1 (i.e. for downregulation).

23

ACCEPTED MANUSCRIPT

Table 2. Biochemicals (100 µM) and genes (in italics, 10 or 100 µM) significantly affected both by BPA and GEN in Atlantic salmon hepatocytes after 48 h

Entrez Gene Name

HMDB

1-stearoyl-2-arachidonoyl-sn-glycero-3-phosphoethanolamine --

Fold change BPA

Fold change GEN

HMDB09003

-1.2

-1.3

SC

Symbol

RI PT

exposure.

HMDB00807

-1.4

-1.4

HMDB02271

1.8

-1.8

-5.3

-4.2

--

5-imidazolepropionic acid

--

Arabonate/xylonate

--

cdkn1b

cyclin dependent kinase inhibitor 1B

”CDKN1B”

2.1

2.2

cyp1a1

cytochrome P450 family 1 subfamily A member 1 ”CYP1A1”

3.1

2.3

D-glucuronic acid

--

HMDB00127

-4.5

-1.6

D-mannose

--

HMDB00169

-1.9

-2.2

D-pantothenic acid

--

HMDB00210

1.8

1.3

Dodecanedioic acid

--

HMDB00623

-3.1

-2.2

Ergothioneine

--

HMDB03045

-2.6

-2.6

esr1

estrogen receptor 1

”ESR1”

3.4

3.4

Gluconic acid-6-phosphate

--

HMDB01316

-2.6

-4.5

--

1.3

1.3

--

1.3

1.4

Glycosyl ceramide (d18:1/23:1. d17:1/24:1) Hydantoin-5-propionate

TE D EP

AC C

Glycosyl ceramide (d16:1/24:1. d18:1/22:1)

M AN U

3-phosphoglycerate

--

HMDB01212

2.2

-1.3

Hypotaurine

--

HMDB00965

-4.3

-8.3

Inosine

--

HMDB00195

-2.0

-1.6

24

insulin induced gene 1

”INSIG1”

2.8

3.5

L-gulonic acid

--

HMDB03290

-4.5

-7.1

Maltose

--

HMDB00163

-10

-12.5

Maltotetraose

--

HMDB01296

-1.3

-1.3

Maltotriose

--

HMDB01262

-7.7

-10

Meglutol

--

HMDB00355

1.6

2.7

N6-succinyladenosine

--

7.7

2.3

N-acetyl-L-glutamate

--

3.5

3.0

N-acetyl-L-histidine

--

HMDB32055

2.7

2.7

N-acetylglucosamine

--

HMDB00215

3.2

3.3

Phosphoenolpyruvate (PEP)

--

HMDB00263

-1.3

-1.5

Prolylhydroxyproline

--

HMDB06695

6.9

3.6

Succinylcarnitine (C4-DC)

--

2.0

2.0

Taurine

--

HMDB00251

-1.4

-2.3

UDP

--

HMDB00295

-3.6

-2.4

UDP-galactose

--

HMDB00302

-2.3

-1.5

UDP-glucuronic acid

--

HMDB00935

-25

-8.3

Uridine

--

HMDB00296

-1.5

-1.3

”VTG1”

18.3

17.2

EP

TE D

M AN U

HMDB01138

AC C

vtg1

RI PT

insig1

SC

ACCEPTED MANUSCRIPT

vitellogenin 1

25

ACCEPTED MANUSCRIPT

Bisphenol A

Score / p-value

Genistein

Endocrine system disorders, organ morphology,

biochemistry, cancer

21

TCA cycle II (Eukaryotic)

2.23E-09

tRNA charging

4.64E-09

Superpathway of methionine degradation Nicotine degradation II

Top upstream regulators

EGFR Afatinib Sirolimus Chlorpyrifos

3.54E-08

Pyrimidine ribonucleotides de novo biosynthesis

1.35E-07

4.26E-08

Estrogen-mediated S-phase entry

1.09E-05

7.85E-08

Alanine degradation III

4.79E-05

Alanine biosynthesis II

4.79E-05

1.56E-23

Afatinib

1.25E-11

7.04E-23

EGFR

4.02E-11

3.32E-16

HADH

2.78E-08

1.20E-08

Sirolimus

1.45E-06

1.63E-07

AC C

Serotonin and melatonin biosynthesis

18

Uridine-5'-phosphate biosynthesis

TE D

pathway

organismal development

EP

Top canonical

M AN U

Drug metabolism, small molecule Top network

Score / p-value

SC

Pathway analysis

RI PT

Table 3. Pathway analysis of significantly affected biochemicals and genes in Atlantic salmon hepatocytes.

26

ACCEPTED MANUSCRIPT

Vinblastine

Zearalenone

1.87E-06

Liver steatosis

2.00E-01 - 2.65E-08

Liver steatosis

Liver necrosis/Cell death

9.87E-02 - 5.85E-08

Liver enlargement

Glutathione depletion in liver

2.64E-02 - 1.17E-06

Liver cholestasis

Liver cholestasis

4.72E-02 - 1.80E-04

Biliary hyperplasia

Liver damage

2.05E-01 - 3.92E-04

Liver inflammation/Hepatitis

SC

RI PT

1.27E-01 - 5.43E-07 2.52E-05 - 2.52E-05 1.98E-02 - 1.75E-04 3.92E-02 - 2.88E-04 1.31E-01 - 1.70E-03

M AN U

Hepatotoxicity

1.71E-08

Entities recognized by the Human Metabolome Database (HMDB) and mapped by IPA (IPA identifiers) were include in pathway analysis. BPA: 99 out of 120

AC C

EP

TE D

biochemicals/genes. GEN: 62 out of 73 biochemicals/genes.

27

ACCEPTED MANUSCRIPT

4. Discussion

This study shows that BPA and GEN, estrogenic compounds known to affect DNA methylation, have relatively similar mode of action in Atlantic salmon liver cells. The study proposes mechanistic

RI PT

biomarkers to be tested in follow-up in vivo studies of the impact of contaminants with similar mode of action found in novel fish feeds. BPA induced global DNA hypomethylation at 1.0 µM, but not at higher exposure concentrations, pointing to a non-linear concentration-response. BPA and GEN

SC

affected transcriptional markers signaling effects on detoxification (BPA: cyp1a1, cyp3a; GEN: cyp1a1), endocrine disruption (BPA: vtg1, esr1, ar; GEN: vtg1, esr1, esr2a), cell survival (BPA: fos,

M AN U

cdkn1b; GEN: cdkn1b), DNA methylation (BPA: dnmt3aa; GEN: dnmt3aa, cbs) and glucose metabolism (BPA and GEN: insig1) in Atlantic salmon hepatocytes. Metabolomic screening showed that BPA and GEN most profoundly affected glucose homeostasis in Atlantic salmon liver cells. Both compounds gave the strongest estrogenic effect at 10 µM, either reflecting a non-linear

TE D

concentration-response pattern or a toxic response disrupting normal cellular functions at 100 µM.

Pathway analysis and predicted upstream regulators suggest that the two chemicals impact many common mechanisms. Although the single metabolite and gene may differ, the pathway analysis

EP

points to many overlapping effects. According to the identified metabolites, both compounds had a profound effect on carbohydrate metabolism. An effect on glucose metabolism was also indicated by

AC C

upregulation of insig1 by both compounds, a gene encoding an endoplasmic reticulum membrane protein that regulates cholesterol metabolism, lipogenesis and glucose homeostasis (GeneCards database). In Atlantic salmon hepatocytes exposed to pirimiphos-methyl (a pesticide found in salmon feeds), a similar upregulation of insig1 transcription was observed (Olsvik et al., 2017). Altered insig1 levels have also been observed in zebrafish (Danio rerio) exposed to bisphenol A (Villeneuve et al., 2012; Saili et al., 2013). Altered glucose metabolism is a well-known effect of endocrine disrupting chemicals. By disrupting endocrine-metabolic pathways these chemicals modulate glucose metabolism and energy balance (Mauvais-Jarvis et al., 2013). A significant drop in glycogen

28

ACCEPTED MANUSCRIPT hydrolysis products suggests that mobilization of glycogen stores in the hepatocytes is affected by BPA and GEN. In combination with lower levels of glucose and glycolytic intermediates, this could indicate an increased glucose demand that cannot be met by glucose uptake and/or glycogen breakdown. These changes could also point to redirection of glucose towards pentose phosphate

RI PT

pathway. The pentose phosphate pathway utilizes glucose 6-phosphate to generate NADPH for biosynthetic reactions and ribose-5-phosphate for nucleotide synthesis (Stanton, 2012). The decrease in pentose sugars and pentose phosphate pathway intermediates may be consistent with a decrease in glucose supply or increased pathway activity to support the demand for NADPH that can be used to

SC

support detoxification reactions (e.g., cytochrome P450, as indicated by cyp1a1 induction, or glutathione reductase, as indicated by altered GSH levels by BPA). Altered TCA cycle II (with 2.3-

M AN U

fold increased citrate levels) was the top affected pathway by BPA exposure, further supporting energy balance disruption as one of the main cellular outcomes.

Disruption of glucose homeostasis and energy generation did however not lead to marked increased

TE D

cytotoxicity at the studied exposure concentrations. With conflicting results for the two cell viability methods, it is difficult to conclude on whether BPA acted cytotoxic at 100 µM. While the MTT assay is a colorimetric assay for assessing cell metabolic activity, the xCELLigence system uses

EP

electrical impedance to quantify cell proliferation, morphology change, and cell adhesion (surface attachment). It has been speculated that early cytotoxicity of BPA is mediated through activation of

AC C

CASP3 in rainbow trout (Oncorhyncus mykiss) (Kaptaner and Kankaya. 2016), however in this study neither BPA nor GEN had any significant effect on casp3b transcription. Cdkn1b, which encodes the cyclin-dependent kinase (Cdk) inhibitor p27 and regulates cell proliferation, cell motility and apoptosis (Chu et al., 2008), was however significantly upregulated by both compounds. Fos, on the other hand, which has also been associated with apoptotic cell death (GeneCards database), was downregulated by the three intermediate BPA concentrations (0.1-10 µM). In a recent in vitro study with Arctic char (Salvelinus alpinus) hepatocytes, Petersen et al. (2017) reported reduced metabolic activity for BPA concentrations above 50 µM and reduced membrane integrity for BPA concentrations above 10 µM. With no response observed with the MTT assay in the current study at

29

ACCEPTED MANUSCRIPT 100 µM BPA, our result indicates that Atlantic salmon hepatocytes are less sensitive to BPA exposure compared to Arctic char hepatocytes. Transcriptional insensitivity to BPA has also been demonstrated in Atlantic salmon kidney cells on oxidative stress markers, with responses only seen at 100 µM (Yazdani et al., 2016). In the current study, altered glucose homeostasis therefore probably

RI PT

reflects an endocrine-metabolic pathway disruption most likely induced by modified estrogen receptor (ER) activity (Kang et al., 2007).

ERs are essential molecules associated with glucose homeostasis and energy balance (Ropero et al.,

SC

2008). In rodents, BPA exposure increases glucose-stimulated insulin release, inducing alterations in glucose homeostasis (Batista et al., 2012). Evidence suggest that this effect relies on BPA binding

M AN U

primarily to the ER alpha (ESR1) in mammals (Alonso-Magdalena et al., 2008; Li et al., 2013). In fish, BPA and similar chemicals also bind to ERs (Villeneuve et al., 2012), and may interfere with insulin-mediated mechanisms (Lombo et al., 2015). Teleosts possess three ER subtypes (Menuet et al., 2002) that are differently expressed and regulated by estrogens (Tohyama et al., 2015; Le Fol et

TE D

al., 2017). In this study, BPA exposure significantly upregulated esr1 at 10 and 100 µM, while GEN upregulated esr1 at 1.0, 10 and 100 µM. Whereas BPA had no impact on esr2a, GEN downregulated esr2a at 100 µM. Both at the mRNA and protein level, many studies have shown impacts of BPA

EP

and GEN on ERs in fish. BPA-induced upregulation of esr1 has been shown in numerous fish studies (e.g. Fu et al., 2007; Huang et al., 2010; Tohyama et al., 2015; Yamaguchi et al., 2015; Sun et al.,

AC C

2014; Qiu et al., 2016). BPA-induced downregulation of esr2a is also commonly seen in fish studies (Huang et al., 2010; Santangeli et al., 2016; Laing et al., 2016). GEN exposure has been shown to bind to and activate ESR1 (Tollefsen et al., 2002; Hawkins and Thomas, 2004; Pinto et al., 2014), ESR2A (Hawkins and Thomas, 2004; Cosnefroy et al., 2012) and ESR2B (Cosnefroy et al., 2012; Pinto et al., 2014) in fish. Furthermore, vtg1, perhaps the most commonly applied biomarker for environmental xenoestrogens in juvenile/male fish (van der Oost et al., 2003), showed a distinct response to BPA and GEN exposure. Vtg1 mirrored esr1 expression after exposure to both BPA and GEN, but showed a larger fold change than esr1. Numerous studies have reported vtg1 upregulation in liver of fish after exposure to BPA (e.g.; Kang et al., 2007; Moens et al., 2007; Villeneuve et al.,

30

ACCEPTED MANUSCRIPT 2012). GEN has also been shown to upregulate vtg1 in fish (Pelissero et al., 1991; Ko et al., 1999; Bennetau-Pelissero et al., 2001; Zhang et al., 2002; Pollack et al., 2003; Scholz and Mayer, 2008; Kausch et al., 2008; Schiller et al., 2013; Bugel et al., 2016). Reduced ar transcription at 100 µM indicates that BPA may behave as an androgen receptor (AR) antagonist, as suggested by some

RI PT

studies (Ekman et al., 2012). This response was not seen for GEN. The current finding indicates a main role for ESR1 in regulating BPA- and GEN-induced effects on carbohydrate metabolism in Atlantic salmon liver cells.

SC

Glutamate levels and its metabolites were affected by the BPA and GEN treatment of hepatocytes. Treatment with BPA resulted in elevated levels of N-acetylglutamate and N-acetylglutamine, while

M AN U

exposure to GEN resulted in higher levels of glutamate and N-acetylglutamate. As for glucose metabolism, BPA is known to affect glutamate metabolism in animals (Cardoso et al., 2010; Zeng et al., 2013; Cabaton et al., 2013). Glutamate is involved in many metabolic reactions, including the TCA cycle (glutaminolysis - entering as alpha-ketoglutarate after deamination), histidine and lysine

TE D

metabolism (downstream catabolite), glutathione biosynthesis (substrate), de novo nucleotide synthesis (glutamine is used for synthesis of carbamoyl phosphate), and urea cycle (carbamoyl phosphate that enters urea cycle is synthesized from glutamine) (GeneCards database). Because

EP

glutamate is central to many biochemical pathways, changes in glutamate levels and/or utilization may affect multiple metabolic processes; also, the reverse is true: changes in metabolic processes,

AC C

e.g. TCA or urea cycle, de novo nucleotide synthesis may result in alteration in glutamate metabolism. Whether these changes have causative links or are secondary to glutamate metabolism would require further investigation.

Both BPA and GEN showed a non-linear concentration-response pattern for some estrogenic mRNA markers in liver cells, with strongest responses at 10 µM for vtg1 and esr1 transcription. A non-linear concentration-response relationship was also indicated by the global DNA methylation data, which showed a hypomethylation response for BPA at 1.0 µM but no response at higher exposure

31

ACCEPTED MANUSCRIPT concentrations. This is in line with previous reports from exposure studies with estrogenic chemicals in fish (Villeneuve et al., 2012;).

BPA is rapidly metabolized in vivo by glucuronidation and sulfation reactions to BPA glucuronide

RI PT

and BPA sulfate in the liver of mammals and fish (Lindholst et al., 2003; Vandenberg et al., 2009). These BPA metabolites are less toxic than the mother compound, and possess little or no estrogenic activity (Vandenberg et al., 2009). Significantly affected BPA metabolites and reduced levels of nucleotide sugar-, amino sugar- and uracil containing pyrimidine metabolites suggests a similar

SC

detoxification mechanism in Atlantic salmon hepatocytes. GEN is also metabolized by the cytochrome P450 system with glucuronidation playing an important role (Breinholt et al., 2003; Smit

M AN U

et al., 2014). Significantly reduced levels of nucleotide sugar-, amino sugar- and uracil containing pyrimidine metabolites in cells exposed to GEN propose an important role for phase II detoxification via glucuronidation reactions in Atlantic salmon hepatocytes. In fish, it has been shown that BPA downregulates cyp1a1 in sea bream (Sparus aurata) (Maradonna et al., 2014), while Bugel et al.

TE D

(2016) observed GEN-mediated downregulation of cyp1a1 in zebrafish. In liver of Atlantic cod (Gadus morhua), we observed a trend toward downregulation of cyp1a after BPA exposure (Olsvik et al., 2009). In this experiment, however, BPA induced cyp1a1 and cyp3a transcription, while GEN

EP

induced cyp1a1. It has been shown that both BPA and GEN can act as AhR agonists in fish (Zhang et al., 2003; Bonefeld-Jorgensen et al., 2007). Receptor-mediated crosstalk between ERs and AhRs has

AC C

been proposed as a model to explain how xenoestrogens such as BPA affect CYP expression in animals, but the mechanisms behind this effect remains poorly known (Vandenberg et al., 2009). Interestingly, BPA exposure also significantly upregulated cyp3a transcription in Atlantic salmon hepatocytes. Although fish CYP3As are considered to be less responsive to contaminants than their mammalian counterparts and show sexual dimorphism of expression patterns, evidence indicate that CYP3As expression is regulated by AHR2 and PXR pathways also in fish (Schlenk et al., 2008; Chang et al., 2013; Gao et al., 2014; Kubota et al., 2013). Taken together, this study suggests that BPA and GEN is partly detoxified by glucuronidation reactions in Atlantic salmon liver cells, possibly following phase I reactions.

32

ACCEPTED MANUSCRIPT

For BPA exposure, an impact on mechanisms linked to DNA methylation was indicated by pathway analysis. The superpathway of methionine degradation, which describes the degradation of Lmethionine, was one of the most significant pathways affected by BPA, based on directional effects

RI PT

on six metabolites. S-adenosylmethionine (SAM), the universal methyl donor for methylation reactions, including DNA methylation, is generated from the methionine and folate cycles (Mentch and Locasale, 2016). In this pathway, L-methionine is first metabolized to S-adenosyl-L-methionine. DNMTs then catalyze the transfer of methyl moieties from S-adenosyl-L-methionine to the cytosine

SC

of CpG dinucleotides (Finkelstein, 1990). SAM was 4.0-fold upregulated by BPA exposure in this study. A similar response on the superpathway of methionine degradation has recently been reported

M AN U

by Strong et al. (2016), studying the effects of BPA on human bone marrow derived mesenchymal stem cells. Contrary to what has previously been reported in fish (Chen et al., 2015; Laing et al., 2016), neither BPA nor GEN exposure induced a downregulation of dnmt1 transcription in Atlantic salmon hepatocytes. Research suggests that expression of dnmt1 is linked to changes in global DNA

TE D

methylation, with inactivation of DNMT1 causing global DNA demethylation (Bestor, 2000). In this study, the demethylating agent 5-AZA was included as a positive control. However, no significant transcriptional responses were seen for 17 genes examined with RT-qPCR (among other dnmt1,

EP

dnmt3a, mat1a2 and cbs) and no significant effect was seen on global DNA methylation in cells exposed to 0.001, 0.01, 0.1, 1 and 10 µM 5-AZA. This might suggest that Atlantic salmon

AC C

hepatocytes are relatively insensitive to demethylating agents, or that promoter targeted methods are required to detect changes in DNA methylation in cultured fish primary liver cells. In ovaries of female zebrafish, it has been shown that the DNA methylation pattern in the promoter region of the dnmt1 gene is associated with BPA exposure concentrations for four CpG sites (Laing et al., 2016). Similar associations were seen in testes of male zebrafish, in which BPA-induced reduced dnmt1 transcription was associated with significant hypermethylation of two CpG sites in the promoter region of the dnmt1 gene (Laing et al., 2016). In general, exposure to BPA and similar chemicals induces genome-wide hypomethylation in fish and other animals (Shugart, 1990; Dolinoy et al., 2007; Bollati and Baccarelli, 2010; Vandegehuchte and Janssen, 2014; Mirbahai et al., 2011). In this

33

ACCEPTED MANUSCRIPT study, however, reduced global DNA methylation at 1.0 µM BPA was not associated with the transcriptional levels of dnmt1 in Atlantic salmon hepatocytes. Rather, both chemicals downregulated dnmt3aa at 10 µM (GEN) and 100 µM (BPA and GEN). The dnmt3aa gene encodes a DNA methyltransferase that is thought to function in de novo methylation, rather than maintenance

RI PT

methylation as for dnmt1 (GeneCards database). Reduced transcription of dnmt3 has previously been reported in zebrafish F2 embryos from parents exposed to BPA (Chen et al., 2015), a response also seen in rat embryos after neonatal exposure to BPA (Doshi et al., 2012). For GEN exposure, the expression pattern of cbs mirrored that of dnmt3aa. The protein encoded by cbs functions in the

SC

folate metabolism pathway and is linked to methylation of genomic DNA (Bao et al., 1998). BPAinduced down-regulation of cbs has previously been reported in mammals (Weng et al., 2010; Ali et

M AN U

al., 2014), and the cbs protein has been proposed as an epigenetic biomarker (Zhao et al., 2012). Taken together, impact on the methionine degradation pathway and dnmt3aa appears as a potential biomarker for BPA exposure in fish liver scells, while dnmt3aa and cbs might be candidate markers

TE D

for GEN exposure.

In conclusion, this study shows that BPA and GEN induce cellular mechanisms controlled by similar upstream regulators in Atlantic salmon hepatocytes. Both compound had most profound effect on

EP

glucose metabolism, likely mediated by binding to ESR1, and also impacted glutamate metabolism. BPA affected metabolites associated with methionine homeostasis and possible the supply of methyl

AC C

groups in the cells. BPA, but not GEN, induced hypomethylation at an intermediate concentration (1.0 µM). BPA had an anti-androgenic effect at 100 µM, a response not seen for GEN. Both phase I (cyp1a1) and phase II (glucuronidation) reactions appear to be involved in the detoxification of BPA and GEN in Atlantic salmon hepatocytes. Potential biomarkers of BPA and GEN exposure in fish primary cells include metabolites associated with glucose and glutamate metabolism, in addition to mRNA markers cyp1a1, vtg1, esr1, esr2a (only GEN), ar (only BPA), dnmt3aa, cbs (only GEN) and cdkn1b.

34

ACCEPTED MANUSCRIPT

Acknowledgment The authors want to thank Hui-shan Tung, Anne Karin Syversen, Synnøve Winterthun and Betty Irgens (NIFES) for technical help. This project was funded by the Norwegian Research Council

AC C

EP

TE D

M AN U

SC

RI PT

(EpiSip project 228877).

35

ACCEPTED MANUSCRIPT

References

Abassi, Y.A., Xi, B., Zhang, W.F., Ye, P.F., Kirstein, S.L., Gaylord, M.R., Feinstein, S.C., Wang, X.B., Xu, X., 2009. Kinetic cell-based morphological screening: Prediction of mechanism of

RI PT

compound action and off-target effects. Chem. Biol. 16, 712-723.

Ali, S., Steinmetz, G., Montillet. G., Perrard, M.H., Loundou, A., Durand, P., Guichaoua, M.R., Prat,

SC

O., 2014. Exposure to low-dose bisphenol A impairs meiosis in the rat seminiferous tubule culture

M AN U

model: A physiotoxicogenomic approach. PLoS ONE 9(9): e106245.

Alonso-Magdalena, P., Ropero, A.B., Carrera, M.P., Cederroth, C.R., Baquie, M., Gauthier, B.R., Nef, S., Stefani, E., Nadal, A., 2008. Pancreatic insulin content regulation by the estrogen receptor ER alpha. PLoS ONE 3(4): e2069.

TE D

Bao, L.M., Vlcek, C., Paces, V., Kraus, J.P., 1998. Identification and tissue distribution of human cystathionine beta-synthase mRNA isoforms. Arch. Biochem. Biophys. 350. 95-103.

EP

Batista, T.M., Alonso-Magdalena, P., Vieira, E., Amaral, M.E.C., Cederroth, C.R., Nef, S., Quesada, I., Carneiro, E.M., Nadal, A., 2012. Short-term treatment with bisphenol-A leads to metabolic

AC C

abnormalities in adult male mice. PLoS ONE 7(3): e33814.

Beischlag, T.V., Luis Morales, J., Hollingshead, B.D., Perdew, G.H., 2008. The aryl hydrocarbon receptor complex and the control of gene expression. Crit. Rev. Eukaryot. Gene Expr. 18. 207-250.

Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate - a practical and powerful approach to multiple testing. J. Roy. Stat. Soc. B Met. 57, 289-300.

36

ACCEPTED MANUSCRIPT Bennetau-Pelissero, C., Breton, B., Bennetau, B., Corraze, G., Le Menn, F., Davail-Cuisset, B., Helou, C., Kaushik, S.J., 2001. Effect of genistein-enriched diets on the endocrine process of gametogenesis and on reproduction efficiency of the rainbow trout Oncorhynchus mykiss. Gen.

RI PT

Comp. Endocrinol. 121. 173-187.

Berntssen, M.H.G., Julshamn, K., Lundebye, A.K., 2010. Chemical contaminants in aquafeeds and Atlantic salmon (Salmo salar) following the use of traditional- versus alternative feed ingredients.

SC

Chemosphere 78. 637-646.

M AN U

Bestor, T.H., 2000. The DNA methyltransferases of mammals. Hum. Mol. Gen. 9. 2395-2402.

Bollati, V., Baccarelli, A., 2010. Environmental epigenetics. Heredity 105(1). 105-112.

Bonefeld-Jorgensen, E.C., Long, M.H., Hofmeister, M.V., Vinggaard, A.M., 2007. Endocrine-

TE D

Disrupting potential of bisphenol A, bisphenol A dimethacrylate, 4-n-nonylphenol, and 4-noctylphenol in vitro: New data and a brief review. Environ. Health Perspect. 115. 69-76.

EP

Breinholt, V.M., Rasmussen, S.E., Brosen, K., Friedberg, T.H., 2003. In vitro metabolism of

AC C

genistein and tangeretin by human and murine cytochrome p450s. Pharmacol. Toxicol. 93, 14-22.

Bugel, S.M., Bonventre, J.A., Tanguay, R.L., 2016. Comparative developmental toxicity of flavonoids using an integrative zebrafish system. Toxicol. Sci. 154, 55-68.

Cabaton, N.J., Canlet, C., Wadia, P.R., Tremblay-Franco, M., Gautier, R., Molina, J., Sonnenschein, C., Cravedi, J.P., Rubin, B.S., Soto, A.M., Zalko, D., 2013. Effects of low doses of bisphenol A on the metabolome of perinatally exposed CD-1 mice. Environ. Health Perspect. 121, 586-593.

37

ACCEPTED MANUSCRIPT Cardoso, N., Pandolfi, M., Ponzo, O., Carbone, S., Szwarcfarb, B., Scacchi, P., Reynoso, R., 2010. Evidence to suggest glutamic acid involvement in bisphenol A effect at the hypothalamic level in prepubertal male rats. Neuroendocrinol. Lett. 31, 512-516.

RI PT

Chang, C.T., Chung, H.Y., Su, H.T., Tseng, H.P., Tzou, W.S., Hu, C.H., 2013. Regulation of zebrafish CYP3A65 transcription by AHR2. Toxicol. Appl. Pharmacol. 270, 174-184.

Chen, J.F., Xiao, Y.Y., Gai, Z.X., Li, R., Zhu, Z.X., Bai, C.L., Tanguay, R.L., Xu, X.J., Huang, C.J.,

SC

Dong, Q.X., 2015. Reproductive toxicity of low level bisphenol A exposures in a two-generation

M AN U

zebrafish assay: Evidence of male-specific effects. Aquat. Toxicol. 169. 204-214.

Chu, I.M., Hengst, L., Slingerland, J.M., 2008. The Cdk inhibitor p27 in human cancer: prognostic potential and relevance to anticancer therapy. Nat. Rev. Cancer 8, 253-267.

TE D

Cosnefroy, A., Brion, F., Maillot-Marechal, E., Porcher, J.M., Pakdel, F., Balaguer, P., Ait-Aissa, S., 2012. Selective activation of zebrafish estrogen receptor subtypes by chemicals by using stable

EP

reporter gene assay developed in a zebrafish liver cell line. Toxicol. Sci. 125, 439-449.

Dolinoy, D.C., Weidman, J.R., Waterland, R.A., Jirtle, R.L., 2006. Maternal genistein alters coat

AC C

color and protects Avy mouse offspring from obesity by modifying the fetal epigenome. Environ. Health Perspect. 114, 567-572.

Dolinoy, D.C., Huang, D., Jirtle, R.L., 2007. Maternal nutrient supplementation counteracts bisphenol A-induced DNA hypomethylation in early development. Proc. Natl. Acad. Sci. USA 104, 13056-13061.

38

ACCEPTED MANUSCRIPT Doshi, T., D'Souza, C., Dighe, V., Vanage, G., 2012. Effect of neonatal exposure on male rats to bisphenol a on the expression of DNA methylation machinery in the postimplantation embryo. J. Biochem. Mol. Toxic. 26, 337-343.

RI PT

Ekman, D.R., Hartig, P.C., Cardon, M., Skelton, D.M., Teng, Q., Durhan, E.J., Jensen, K.M., Kahl, M.D., Villeneuve, D.L., Gray, L.E., Collette, T.W., Ankley, G.T., 2012. Metabolite profiling and a transcriptional activation assay provide direct evidence of androgen receptor antagonism by

SC

bisphenol A in fish. Environ. Sci. Technol. 46, 9673-9680.

Fan, S., Meng, Q., Auborn, K., Carter, T., Rosen, E.M., 2006. BRCA1 and BRCA2 as molecular

M AN U

targets for phytochemicals indole-3-carbinol and genistein in breast and prostate cancer cells. Brit. J. Cancer 94, 407-426.

TE D

Finkelstein, J.D., 1990. Methionine Metabolism in Mammals. J. Nutr. Biochem. 1, 228-237.

Fu, K.Y., Chen, C.Y., Chang, W.M., 2007. Application of a yeast estrogen screen in non-biomarker species Varicorhinus barbatulus fish with two estrogen receptor subtypes to assess xenoestrogens.

EP

Toxicol. In Vitro 21, 604-612.

AC C

Gao, J.C., Zhang, Y.Y., Yang, Y.P., Yuan, C., Qin, F., Liu, S.Z., Zheng, Y., Wang, Z.Z., 2014. Molecular characterization of PXR and two sulfotransferases and hepatic transcripts of PX, two sulfotransferases and CYP3A responsive to bisphenol A in rare minnow Gobiocypris rarus. Mol. Biol. Rep. 41, 7153-7165.

GeneCards database [http://www.genecards.org], assessed May 5th, 2017.

Hansen, A.C., Waagbo, R., Hemre, G.I., 2015. New B vitamin recommendations in fish when fed plant-based diets. Aquacult. Nutr. 21, 507-527.

39

ACCEPTED MANUSCRIPT

Hawkins, M.B., Thomas, P., 2004. The unusual binding properties of the third distinct teleost estrogen receptor subtype ER beta a are accompanied by highly conserved amino acid changes in the

RI PT

ligand binding domain. Endocrinol. 145, 2968-2977.

Hemre, G.I., Lock, E.J., Olsvik, P.A., Hamre, K., Espe, M., Torstensen, B.E., Silva, J., Hansen, A.C., Waagbo, R., Johansen, J.S., Sanden, M., Sissener, N.H., 2016. Atlantic salmon (Salmo salar) require increased dietary levels of B-vitamins when fed diets with high inclusion of plant based ingredients.

SC

PeerJ 4, e2493.

M AN U

Hites, R.A., Foran, J.A., Carpenter, D.O., Hamilton, M.C., Knuth, B.A., Schwager, S.J., 2004. Global assessment of organic contaminants in farmed salmon. Science 303, 226-229.

Huang, W.R., Zhang, Y., Jia, X.P., Ma, X.L., Li, S.S., Liu, Y., Zhu, P., Lu, D.Q., Zhao, H.H., Luo,

TE D

W.N., Yi, S.B., Liu, X.C., Lin, H.R., 2010. Distinct expression of three estrogen receptors in response to bisphenol A and nonylphenol in male Nile tilapias (Oreochromis niloticus). Fish Physiol.

EP

Biochem. 36, 237-249.

ISO, 2009. ISO 10993-5:2009 - Biological evaluation of medical devices Part 5. Tests for in vitro

AC C

cytotoxicity. Available: . Accessed 15 March 2016.

Kang, J.H., Aasi, D., Katayama, Y., 2007. Bisphenol A in the aquatic environment and its endocrinedisruptive effects on aquatic organisms. Crit. Rev. Toxicol. 37, 607-625.

Kaptaner, B., Kankaya, E., 2016. Caspase-3 activation in cytotoxicity of isolated rainbow trout (Oncorhyncus mykiss) hepatocytes induced by bisphenol A. Fres. Environ. Bull. 25, 1167-1174.

40

ACCEPTED MANUSCRIPT Kausch, U., Alberti, M., Haindl, S., Budczies, J., Hock, B., 2008. Biomarkers for exposure to estrogenic compounds: Gene expression analysis in zebrafish (Danio rerio). Environ. Toxicol. 23, 15-24.

RI PT

Kennedy, C.J., Osachoff, H.L., Shelley, L.K., 2014. Estrogenic endocrine disrupting chemicals in fish. In: Tierney, K.B., Farrell, A.P., Brauner, C.J. (Eds.). Organic chemical toxicology of fishes. Academic Press. Amsterdam. pp. 257-307.

SC

Kim, D.J., Seok, S.H., Baek, M.W., Lee, H.Y., Na, Y.R., Park, S.H., Lee, H.K., Dutta, N.K., Kawakami, K., Park, J.H., 2009. Developmental toxicity and brain aromatase induction by high

M AN U

genistein concentrations in zebrafish embryos. Toxicol. Mech. Method 19, 251-256.

Ko, K., Malison, J.A., Reed, J.D., 1999. Effect of genistein on the growth and reproductive function

TE D

of male and female yellow perch Perca flavescens. J. World Aquacult. Soc. 30, 73-79.

Krishnan, A.V., Stathis, P., Permuth, S.F., Tokes, L., Feldman, D., 1993. Bisphenol-A - an estrogenic

EP

substance is released from polycarbonate flasks during autoclaving. Endocrinol. 132, 2279-2286.

Krøvel, A.V., Søfteland, L., Torstensen, B.E., Olsvik, P.A., 2010. Endosulfan in vitro toxicity in

AC C

Atlantic salmon hepatocytes obtained from fish fed either fish oil or vegetable oil. Comp. Biochem. Physiol. C. 151(2), 175-186.

Kubota, A., Bainy, A.C.D., Woodin, B.R., Goldstone, J.V., Stegeman, J.J., 2013. The cytochrome P450 2AA gene cluster in zebrafish (Danio rerio): Expression of CYP2AA1 and CYP2AA2 and response to phenobarbital-type inducers. Toxicol. Appl. Pharmacol. 272, 172-179.

41

ACCEPTED MANUSCRIPT Le Fol, V., Ait-Aissa, S., Sonavane, M., Porcher, J.M., Balaguer, P., Cravedi, J.P., Zalko, D., Brion, F., 2017. In vitro and in vivo estrogenic activity of BPA, BPF and BPS in zebrafish-specific assays. Ecotoxicol. Environ. Saf. 142, 150-156.

RI PT

Li, Y., Luh, C.J., Burns, K.A., Arao, Y., Jiang, Z.L., Teng, C.T., Tice, R.R., Korach, K.S., 2013. Endocrine-disrupting chemicals (EDCs): In vitro mechanism of estrogenic activation and differential effects on ER target genes. Environ. Health Perspect. 121, 459-466.

SC

Laing, L.V., Viana, J., Dempster, E.L., Trznadel, M., Trunkfield, L.A., Uren Webster, T.M., van Aerle, R., Paull, G.C., Wilson, R.J., Mill, J., Santos, E.M., 2016. Bisphenol A causes reproductive

M AN U

toxicity, decreases dnmt1 transcription, and reduces global DNA methylation in breeding zebrafish (Danio rerio). Epigenetics-US 11, 526-538.

Lindholst, C., Wynne, P.M., Marriott, P., Pedersen, S.N., Bjerregaard, P., 2003. Metabolism of

TE D

bisphenol A in zebrafish (Danio rerio) and rainbow trout (Oncorhynchus mykiss) in relation to estrogenic response. Comp. Biochem. Physiol. C. 135, 169-177.

EP

Lombo, M., Fernandez-Diez, C., Gonzalez-Rojo, S., Navarro, C., Robles, V., Herraez, M.P., 2015. Transgenerational inheritance of heart disorders caused by paternal bisphenol A exposure. Environ.

AC C

Poll. 206, 667-678.

Maradonna, F., Nozzi, V., Valle, L.D., Traversi, I., Gioacchini, G., Benato, F., Colletti, E., Gallo, P., Di Marco Pisciottano, I., Mita, D.G., Hardiman, G., Mandich, A., Carnevali, O., 2014. A developmental hepatotoxicity study of dietary bisphenol A in Sparus aurata juveniles. Comp. Biochem. Physiol. C. 166, 1-13.

Mauvais-Jarvis, F., Clegg, D.J., Hevener, A.L., 2013. The role of estrogens in control of energy balance and glucose homeostasis. Endocr. Rev. 34, 309-338.

42

ACCEPTED MANUSCRIPT

Mentch, S.J., Locasale, J.W., 2016. One-carbon metabolism and epigenetics: understanding the specificity. Ann. Ny. Acad. Sci. 1363, 91-98.

RI PT

Menuet, A., Pellegrini, E., Anglade, I., Blaise, O., Laudet, V., Kah, O., Pakdel, F., 2002. Molecular characterization of three estrogen receptor forms in zebrafish: Binding characteristics. transactivation properties. and tissue distributions. Biol. Reprod. 66, 1881-1892.

SC

Mirbahai., L., Yin, G.L., Bignell, J.P., Li, N., Williams, T.D., Chipman, J.K., 2011. DNA

M AN U

methylation in liver tumorigenesis in fish from the environment. Epigenetics-US 6, 1319-1333.

Moens, L.N., van der Ven, K., Van Remortel, R., Del-Favero, J., De Coen, W.M., 2007. Gene expression analysis of Estrogenic compounds in the liver of common carp (Cyprinus carpio) using a

TE D

custom cDNA Microarray. J. Biochem. Mol. Toxic. 21, 299-311.

Morito, K., Hirose, T., Kinjo, J., Hirakawa, T., Okawa, M., Nohara, T., Ogawa, S., Inoue, S., Muramatsu, M., Masamune, Y., 2001. Interaction of phytoestrogens with estrogen receptors alpha

EP

and beta. Biol. Pharm. Bull. 24, 351-356.

AC C

Nacher-Mestre, J., Serrano, R., Portoles, T., Berntssen, M.H., Perez-Sanchez, J., Hernandez, F., 2014. Screening of pesticides and polycyclic aromatic hydrocarbons in feeds and fish tissues by gas chromatography coupled to high-resolution mass spectrometry using atmospheric pressure chemical ionization. J. Agric. Food Chem. 62, 2165-2174.

Nøstbakken, O.J., Hove, H.T., Duinker, A., Lundebye, A.K., Berntssen, M.H., Hannisdal, R., Lunestad, B.T., Maage, A., Madsen, L., Torstensen, B.E., Julshamn, K., 2015. Contaminant levels in Norwegian farmed Atlantic salmon (Salmo salar) in the 13-year period from 1999 to 2011. Environ Int 74, 274-280.

43

ACCEPTED MANUSCRIPT

Olsvik, P.A., Lie, K.K., Sturve, J., Hasselberg, L., Andersen, O.K., 2009. Transcriptional effects of nonylphenol, bisphenol A and PBDE-47 in liver of juvenile Atlantic cod (Gadus morhua).

RI PT

Chemosphere 75, 360-367.

Olsvik, P.A., Berntssen, M.H., Søfteland, L., 2015a. Modifying effects of vitamin E on chlorpyrifos toxicity in Atlantic salmon. PLoS ONE 10, e0119250.

SC

Olsvik, P.A., Samuelsen, O.B., Agnalt, A.L., Lunestad, B.T., 2015b. Transcriptional responses to

M AN U

teflubenzuron exposure in European lobster (Homarus gammarus). Aquat. Toxicol. 167, 143-156.

Olsvik, P.A., Berntssen, M.H., Søfteland, L., 2017. In vitro toxicity of pirimiphos-methyl in Atlantic salmon hepatocytes. Toxicol. In Vitro 39, 1-14.

TE D

Patisaul, H.B., Adewale, H.B., 2009. Long-term effects of environmental endocrine disruptors on reproductive physiology and behavior. Front. Behav. Neurosci. 3, 10.

EP

Petersen, K., Hultman, M.T., Tollefsen, K.T., 2017. Primary hepatocytes from Arctic char (Salvelinus alpinus) as a relevant Arctic in vitro model for screening contaminants and environmental

AC C

extracts. Aquat. Toxicol. 187, 141–152.

Pelissero, C., Bennetau, B., Babin, P., Lemenn, F., Dunogues, J., 1991. The estrogenic activity of certain phytoestrogens in the Siberian sturgeon Acipenser-Baeri. J. Steroid Biochem. 38, 293-299.

Pinto, C., Grimaldi, M., Boulahtouf, A., Pakdel, F., Brion, F., Ait-Aissa, S., Cavailles, V., Bourguet, W., Gustafsson, J.A., Bondesson, M., Balaguer, P., 2014. Selectivity of natural. synthetic and environmental estrogens for zebrafish estrogen receptors. Toxicol. Appl. Pharmacol. 280, 60-69.

44

ACCEPTED MANUSCRIPT Pollack, S.J., Ottinger, M.A., Sullivan, C.V., Woods, L.C., 2003. The effects of the soy isoflavone genistein on the reproductive development of striped bass. N. Am. J. Aquacult. 65, 226-234.

Qiu, W.H., Zhao, Y.L., Yang, M., Farajzadeh, M., Pan, C.Y., Wayne, N.L., 2016. Actions of

RI PT

bisphenol A and bisphenol S on the reproductive neuroendocrine system during early development in zebrafish. Endocrinol. 157, 636-647.

performance liquid chromatography. Methods 27, 156-161.

SC

Ramsahoye, B.H., 2002. Measurement of genome wide DNA methylation by reversed-phase high-

M AN U

Ropero, A.B., Alonso-Magdalena, P., Quesada, I., Nadal, A., 2008. The role of estrogen receptors in the control of energy and glucose homeostasis. Steroids 73, 874-879.

Saili, K.A., Tilton, S.C., Waters, K.M., Tanguay, R.L., 2013. Global gene expression analysis reveals

TE D

pathway differences between teratogenic and non-teratogenic exposure concentrations of bisphenol A and 17β-estradiol in embryonic zebrafish. Reprod. Toxicol. (Elmsford, N.Y.) 38, 89-101.

EP

Santangeli, S., Maradonna, F., Gioacchini, G., Cobellis, G., Piccinetti, C.C., Dalla Valle, L., Carnevali, O., 2016. BPA-induced deregulation of epigenetic patterns: Effects on female zebrafish

AC C

reproduction. Sci. Rep. UK 6:21982.

Schiller, V., Wichmann, A., Kriehuber, R., Muth-Kohne, E., Giesy, J.P., Hecker, M., Fenske, M., 2013. Studying the effects of genistein on gene expression of fish embryos as an alternative testing approach for endocrine disruption. Comp. Biochem. Physiol. C. 157, 41-53.

Schlenk, D., Celander, M., Gallagher, E.P., George, S., James, M., Kullman, S.W., Hurk, P.V.D., Willett, K., 2008. Biotransformation in fishes. in: Guiulio, R.T., Hinton, D.E. (Eds.). The Toxicology of Fishes. CRC Press. Boca Raton. pp. 153–234.

45

ACCEPTED MANUSCRIPT

Scholz, S., Mayer, I., 2008. Molecular biomarkers of endocrine disruption in small model fish. Mol. Cell. Endocrinol. 293, 57-70.

RI PT

Shugart, L.R., 1990. 5-Methyl deoxycytidine content of DNA from bluegill sunfish (Lepomismacrochirus) exposed to benzo[a]Pyrene. Environ. Toxicol. Chem. 9, 205-208.

Skjaerven, K.H., Hamre, K., Penglase, S., Finn, R.N., Olsvik, P.A., 2014. Thermal stress alters

SC

expression of genes involved in one carbon and DNA methylation pathways in Atlantic cod embryos.

M AN U

Comp. Biochem. Physiol. A. 173C, 17-27.

Smit, S., Szymanska, E., Kunz, I., Roldan, V.G., van Tilborg, M.W.E.M., Weber, P., Prudence, K., van der Kloet, F.M., van Duynhoven, J.P.M., Smilde, A.K., de Vos, R.C.H., Bendik, I., 2014. Nutrikinetic modeling reveals order of genistein phase II metabolites appearance in human plasma.

TE D

Mol. Nutr. Food Res. 58, 2111-2121.

Søfteland, L., Eide, I., Olsvik, P.A., 2009. Factorial design applied for multiple endpoint toxicity

EP

evaluation in Atlantic salmon (Salmo salar L.) hepatocytes. Toxicol. In Vitro 23, 1455-1464.

AC C

Søfteland, L., Kirwan, J.A., Hori, T.S.F., Størseth, T.R., Sommer, U., Berntssen, M.H.G., Viant, M.R., Rise, M.L., Waagbø, R., Torstensen, B.E., Booman, M., Olsvik, P.A., 2014. Toxicological effect of single contaminants and contaminant mixtures associated with plant ingredients in novel salmon feeds. Food Chem. Toxicol. 73, 157–174.

Søfteland, L., Berntssen, M.H.G., Kirwan, J.A., Størseth, T.R., Viant, M.R., Torstensen, B.E., Waagbø, R., Olsvik, P.A., 2016. Omega-3 and alpha-tocopherol provide more protection against contaminants in novel feeds for Atlantic salmon (Salmo salar L.) than omega-6 and gamma tocopherol. Toxicol. Rep. 3, 211–224.

46

ACCEPTED MANUSCRIPT

Stanton, R.C., 2012. Glucose-6-phosphate dehydrogenase, NADPH, and cell survival. IUBMB Life 64, 362-369.

RI PT

Staples, C.A., Dorn, P.B., Klecka, G.M., O'Block, S.T., Harris, L.R., 1998. A review of the environmental fate, effects, and exposures of bisphenol A. Chemosphere 36, 2149-2173.

Strong, A.L., Miller, D.F.B., Buechlein, A.M., Fang, F., Glowacki, J., McLachlan, J.A., Nephew,

SC

K.P., Burow, M.E., Bunnell, B.A., 2016. Bisphenol A alters the self-renewal and differentiation

M AN U

capacity of human bone-marrow- derived mesenchymal stem cells. Endocr. Disrupt. 4:1, e1200344.

Sumpter, J.P., Jobling, S., 1995. Vitellogenesis as a biomarker for estrogenic contamination of the aquatic environment. Environ. Health Perspect. 103 Suppl. 7, 173-178.

TE D

Sun, L.W., Lin, X., Jin, R., Peng, T., Peng, Z.H., Fu, Z.W., 2014. Toxic effects of bisphenol A on early life stages of Japanese medaka (Oryzias latipes). Bull. Environ. Contam. Tox. 93, 222-227.

EP

Tohyama, S., Miyagawa, S., Lange, A., Ogino, Y., Mizutani, T., Tatarazako, N., Katsu, Y., Ihara, M., Tanaka, H., Ishibashi, H., Kobayashi, T., Tyler, C.R., Iguchi, T., 2015. Understanding the molecular

AC C

basis for differences in responses of fish estrogen receptor subtypes to environmental estrogens. Environ. Sci. Technol. 49, 7439-7447.

Tollefsen, K.E., Mathisen, R., Stenersen, J., 2002. Estrogen mimics bind with similar affinity and specificity to the hepatic estrogen receptor in Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss). Gen. Comp. Endocrinol. 126, 14-22.

Vandegehuchte, M.B., Janssen, C.R., 2014. Epigenetics in an ecotoxicological context. Mutat. Res. Gen. Tox. En. 764, 36-45.

47

ACCEPTED MANUSCRIPT

Vandesompele, J., De Preter, K., Pattyn, F., Poppe, B., Van Roy, N., De Paepe, A., Speleman, F., 2002. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of

RI PT

multiple internal control genes. Genome Biol. 3, RESEARCH0034.

Vandenberg, L.N., Maffini, M.V., Sonnenschein, C., Rubin, B.S., Soto, A.M., 2009. Bisphenol-A and the great divide: A review of controversies in the field of endocrine disruption. Endocr. Rev. 30,

SC

75-95.

van der Oost, R., Beyer, J., Vermeulen, N.P.E., 2003. Fish bioaccumulation and biomarkers in

M AN U

environmental risk assessment: a review. Environ. Toxicol. Pharmacol. 13, 57-149.

Villeneuve, D.L., Garcia-Reyero, N., Escalon, B.L., Jensen, K.M., Cavallin, J.E., Makynen, E.A., Durhan, E.J., Kahl, M.D., Thomas, L.M., Perkins, E.J., Ankley, G.T., 2012. Ecotoxicogenomics to

51-59.

TE D

support ecological risk assessment: A case study with bisphenol A in fish. Environ. Sci. Technol. 46,

EP

Weng, Y.I., Hsu, P.Y., Liyanarachchi, S., Liu, J., Deatherage, D.E., Huang, Y.W., Zuo, T., Rodriguez, B., Lin, C.H., Cheng, A.L., Huang, T.H.M., 2010. Epigenetic influences of low-dose

AC C

bisphenol A in primary human breast epithelial cells. Toxicol. Appl. Pharmacol. 248, 111-121.

Yamaguchi, A., Ishibashi, H., Arizono, K., Tominaga, N., 2015. In vivo and in silico analyses of estrogenic potential of bisphenol analogs in medaka (Oryzias latipes) and common carp (Cyprinus carpio). Ecotoxicol. Environ. Saf. 120, 198-205.

Yazdani, M., Andresen, A.M.S., Gjoen, T., 2016. Short-term effect of bisphenol-a on oxidative stress responses in Atlantic salmon kidney cell line: a transcriptional study. Toxicol. Mech. Method. 26, 295-300.

48

ACCEPTED MANUSCRIPT

Zeng, J., Kuang, H., Hu, C.X., Shi, X.Z., Yan, M., Xu, L.G., Wang, L.B., Xu, C.L., Xu, G.W., 2013. Effect of bisphenol a on rat metabolic profiling studied by using capillary electrophoresis time-of-

RI PT

flight mass spectrometry. Environ. Sci. Technol. 47, 7457-7465.

Zhang, L.L., Khan, I.A., Foran, C.M., 2002. Characterization of the estrogenic response to genistein in Japanese medaka (Oryzias latipes). Comp. Biochem. Physiol. C. 132, 203-211.

SC

Zhang, S., Qin, C.H., Safe, S.H., 2003. Flavonoids as aryl hydrocarbon receptor agonists/antagonists:

M AN U

Effects of structure and cell context. Environ. Health Perspect. 111, 1877-1882.

Zhao, H., Li, Q.S., Wang, J., Su, X.W., Ng, K.M., Qiu, T., Shan, L., Ling, Y., Wang, L.F., Cai, J.Q., Ying, J.M., 2012. Frequent epigenetic silencing of the folate-metabolising gene cystathionine-beta-

AC C

EP

TE D

synthase in gastrointestinal cancer. PLoS ONE 7(11): e49683.

49

ACCEPTED MANUSCRIPT

Supplementary Files

Supplementary File 1. Cytotoxicity and global DNA methylation. Fig. S1: Cytotoxicity of BPA and GEN in Atlantic salmon hepatocytes as determined by the MTT

hepatocytes exposed to BPA and GEN.

SC

Supplementary File 2. Metabolomic screening.

RI PT

assay and the xCELLigence system. Fig. S2: Global DNA methylation in Atlantic salmon

Biochemicals significantly affected by exposure to 100 µM BPA and GEN for 48 h in Atlantic

AC C

EP

TE D

M AN U

salmon hepatocytes.

50

ACCEPTED MANUSCRIPT

Highlights Atlantic salmon hepatocytes were exposed to bisphenol A and genistein for 48 h

BPA and GEN both upregulated esr1, vtg1 and cyp1a transcription

RI PT

Endpoints: cytotoxicity, DNA methylation, targeted transcriptomics and metabolomics

Metabolomics suggests main effects on glucose homeostasis and energy generation

AC C

EP

TE D

M AN U

SC

BPA affected methionine degradation; GEN uridine and pyrimidine biosynthesis