Abstract

   The aim of this study was to elucidate how vitamin E (alpha tocopherol)
   may ameliorate the toxicity of the pesticide chlorpyrifos in Atlantic
   salmon. Freshly isolated hepatocytes were exposed to vitamin E,
   chlorpyrifos or a combination of vitamin E and chlorpyrifos (all 100
   μM). Transcriptomics (RNA-seq) and metabolomics were used to screen for
   effects of vitamin E and chlorpyrifos. By introducing vitamin E, the
   number of upregulated transcripts induced by chlorpyrifos exposure was
   reduced from 941 to 626, while the number of downregulated transcripts
   was reduced from 901 to 742 compared to the control. Adding only
   vitamin E had no effect on the transcriptome. Jak-STAT signaling was
   the most significantly affected pathway by chlorpyrifos treatment
   according to the transcriptomics data. The metabolomics data showed
   that accumulation of multiple long chain fatty acids and dipeptides and
   amino acids in chlorpyrifos treated cells was partially alleviated by
   vitamin E treatment. Significant interaction effects between
   chlorpyrifos and vitamin E were seen for 15 metabolites, including 12
   dipeptides. The antioxidant had relatively modest effects on
   chlorpyrifos-induced oxidative stress. By combining the two data sets,
   the study suggests that vitamin E supplementation prevents uptake and
   accumulation of fatty acids, and counteracts inhibited carbohydrate
   metabolism. Overall, this study shows that vitamin E only to a moderate
   degree modifies chlorpyrifos toxicity in Atlantic salmon liver cells.

Introduction

   Several nutrients have modifying effects on the toxicity of
   contaminants. Interactions between nutrients and contaminants can
   enhance the protection against negative effects of unwanted compounds.
   For example, vitamin E (tocopherols), flavonoids, and fatty acids amend
   the toxicity of polycyclic aromatic hydrocarbons (PAHs) and pesticides
   [[27]1,[28]2,[29]3]. The mechanisms underlying such interactions are
   however often not very well characterized.

   Interactions between nutrients and contaminants are of particular
   interest in fish, and especially in farmed fish such as Atlantic
   salmon, with the recent replacement of fish-based feed ingredients with
   plant-based feed ingredients. Due to overfishing and reduced global
   stocks of fish used to produce commercial feeds [[30]4], the farming
   industry today uses salmon diets >60% of the fish oil replaced with
   plant oil, and with decreasing fishmeal inclusion levels [[31]5].
   Dietary fatty acid composition can affect tissue fatty acids in
   Atlantic salmon in a tissue-dependent pattern [[32]6], especially in
   liver and white muscle [[33]7], and compromise the immune system of the
   fish [[34]8]. In addition, while fish-based oils are rich in
   alpha-tocopherol, plant-based feeds may contain more gamma-tocopherol.
   These forms of vitamin E, which have antioxidant and anti-inflammatory
   properties, may have different abilities to protect the fish against
   the effects of contaminants.

   With increasing inclusion of plant-based ingredients, feeds for
   Atlantic salmon may become contaminated with agricultural pesticides.
   Recently, wide-scale screening of fish feeds for contaminants has
   identified, among others, the insecticides chlorpyrifos-methyl and
   pirmiphos-methyl as possible threats for farmed salmon [[35]9]. New
   plant-based feeds thus not only alter the dietary balance of essential
   nutrients and change the nutritional composition of the fish, but also
   introduce contaminants not normally associated with salmon farming. The
   organophosphate pesticide chlorpyrifos is a broad-spectrum insecticide
   used to kill a wide variety of insects [[36]10]. It remains one of the
   most widely used agricultural organophosphate insecticides, and is
   currently in use in more than 100 countries worldwide
   [[37]11,[38]12,[39]13]. Chlorpyrifos is highly toxic to aquatic
   organisms including fish. It bioaccumulates in fish and has a 96-hour
   LC50 toxicity value in rainbow trout (Oncorhynchus mykiss) between 7.1
   and 51 μg/L, depending on water temperature [[40]14]. Chlorpyrifos has
   at least three main modes of action in mammals. It inhibits the enzyme
   acetylcholinesterase (AChE), causes oxidative stress and endocrine
   disruption [[41]15]. In both mammals and fish, AChE inhibition is the
   main effect of chlorpyrifos exposure [[42]16]. The main detoxification
   system is via the cytochrome P450 enzyme system [[43]17]. Chlorpyrifos
   is completely metabolized to chlorpyrifos oxon and then to
   3,5,6-trichloro-2-pyridinol (TCP) in the mammalian liver by cytochrome
   P450 system [[44]18].

   The aim of this study was to evaluate whether vitamin E (alpha
   tocopherol) amends the toxicity of chlorpyrifos in Atlantic salmon.
   Vitamin E has been reported to be partially protective against
   chlorpyrifos in animal models [[45]19,[46]20,[47]21]. Freshly isolated
   primary hepatocytes were used as a model, and cells were either kept as
   control or treated with 100 μM alpha tocopherol, 100 μM chlorpyrifos or
   a combination of vitamin E and chlorpyrifos for 48 hours. Previous
   experiments have shown the chosen chlorpyrifos concentration to be
   non-cytotoxic to Atlantic salmon hepatocytes, but still high enough to
   induce marked transcriptional responses [[48]22]. Similar experiments
   have been conducted with vitamin E, providing the rationale for using
   the selected concentration of alpha tocopherol in the current study
   (unpublished data). Direct sequencing (RNA-seq) and metabolite
   profiling were used to screen for interactive effects between vitamin E
   and chlorpyrifos. To evaluate whether transcriptional profiling can be
   used to predict the metabolite outcome in the cells, we compared the
   significantly affected KEGG pathways, as identified with
   transcriptional profiling, with the predicted cellular responses based
   on affected metabolites.

Materials and Methods

Fish sampling and cell harvesting

   Fish sacrifice was conducted by the authors and approved by the
   Norwegian Animal Research Authority (NARA) via NIFES' Animal Care and
   Use Committee. Juvenile Atlantic salmon (Salmo salar) were obtained and
   kept at the animal holding facility at Industrilaboratoriet (ILAB),
   Bergen, Norway. The fish were fed once a day a special feed produced
   without synthetic antioxidants and with low levels of contaminants,
   delivered by EWOS, Norway (Spirit 400-50A HH, 6.0 mm). All glassware,
   instruments and solutions were autoclaved prior to liver perfusion.
   Hepatocytes were isolated from 6 male Atlantic salmon (mean±SEM: 555±20
   g) with a two-step perfusion method earlier described by Søfteland et
   al. [[49]23]. The fish were sacrificed by terminal anaesthetization
   with tricaine methanesulfonate (MS-222) (200 mg/l). Harvesting of cells
   was conducted in agreement with and approved by national legislation.
   The final cell pellet was resuspended in L-15 medium containing 10%
   Fish serum (FS) from salmon (Nordic BioSite, Oslo, Norway), 1% glutamax
   (Invitrogen, Norway) and 1% penicillin-streptomycin-amphotericin (10000
   units/ml potassium penicillin 10000 mcg/ml steptomycin sulfate and 25
   μg/ml amphotericin B.) (Lonzo, Medprobe, Oslo, Norway). The Trypan Blue
   exclusion method was performed in accordance with the manufacturer’s
   protocol (Lonzo, Medprobe, Oslo, Norway) and was used to determine
   cells viability. The cell suspensions were plated on 2 μg/cm^2 laminin
   (Sigma-Aldrich, Oslo, Norway) coated culture plates (TPP, Trasadingen,
   Switzerland) and the hepatocytes were kept at 10°C in a sterile
   incubator without additional O[2]/CO[2] (Sanyo, CFC FREE, Etten Leur,
   Netherland). The following cell densities were used; 7.2×10^6 cells per
   well in 6-well plates (in 3 ml complete L-15 medium for transcriptional
   and metabolite profiling) and 0.2×10^6 cells per well in xCELLigence
   96-well plates (in 0.2 ml complete L-15 medium for cytotoxicity
   screening).

Exposure experiments

   The cells were cultured for 36–40 h prior to chemical exposure with
   exchange of medium (containing 10% FS) after 18–20 hours. For the
   exposure experiments, cells were treated with alpha tocopherol (100
   μM), chlorpyrifos (100 μM) or a combination of alpha tocopherol (100
   μM) and chlorpyrifos (100 μM) and harvested after 48 hours exposure.
   [50]Table 1 shows an overview of the number of samples collected for
   the two experiments. Chlorpyrifos was dissolved in DMSO. An equal
   amount of DMSO was used in all four experimental groups. Alpha
   tocopherol and chlorpyrifos were obtained from Sigma (Sigma-Aldrich,
   Oslo, Norway). The cells were exposed in triplicate wells using 6-well
   culture plates for the transcriptional and metabolite profiling, and in
   single 96-wells culture for the xCELLigence cytotoxicity screening. The
   exposure medium contained 1% FS. The exposure medium was exchanged with
   new medium after 18–20 hours and the chemical exposure was sustained
   for another 24 hours.

Table 1. Number of samples used for the various analytical methods.

   Group RNA-seq (cells) Metabolomics (cells) Metabolomics exposure medium
   Description
   Control 6 6 1 Control (w/DMSO)
   Alpha tocopherol (100 μM) 6 6 1 Vitamin E (Vit E) exposed (w/DMSO)
   Chlorpyrifos (100 μM) 6 6 1 Chlorpyrifos (CPF) exposed (w/DMSO)
   Chlorpyrifos (100 μM) + Alpha tocopherol (100 μM) 6 6 1
   Chlorpyrifos/vitamin E exposed (w/DMSO)
   [51]Open in a new tab

Cytotoxicity screening

   Impedance-based real time detection of cellular viability was conducted
   using the xCELLigence system (Real-Time Cell Analyzer RTCA-SP, ACEA
   Biosciences, San Diego, USA), described in detail by Abassi et al.
   [[52]24]. Recording of cell index (CI) values and normalization was
   performed using the RTCA Software version 1.2.1. Primary hepatocyte
   cells were evenly distributed to 96-well E plates. Each well contained
   about 0.2 million cells. Coating and cell density optimization was
   ensured by preliminary experiments. The cells were allowed to attach to
   the 96-well E plates at room temperature (30 min) before being inserted
   in the cell incubator for continuous impedance recording. The real time
   cell monitoring was conducted at 10°C in an incubator without
   additional O[2]/CO[2] (Sanyo, CFC FREE, Etten Leur, Netherland), using
   the RTCA single plate xCELLigence platform. The data was collected with
   intervals of 2 min after compound exposure for 12 hours, and then every
   15 min for 60 hours. 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 about one hour before the
   final exchange of exposure medium, or after about 38 hours of exposure:
   CI[(normalized)] = CI[time x]/CI[norm time] (termed “normalized cell
   index” or NCI). The NCI values were calculated from cells obtained from
   6 different male fish (n = 6). Determination of cytotoxic effect was
   done according to the International standardized test for in vitro
   cytotoxicity ISO 10993-5:2009 [[53]25].

RNA isolation

   Hepatocyte cells from Atlantic salmon were treated with lysis buffer
   before homogenization with the Precellys 24 homogenizer by using
   ceramic beads CK28 (Bertin Technologies, Montigny-le-Bretonneux,
   France). The RNeasy Plus mini kit (Qiagen, Crawley, UK) was used to
   extract total RNA according to the manufacturer’s protocol. The RNA was
   eluted in 30 μl RNase-free MilliQ H[2]O and stored at −80°C before
   further processing. RNA quality and integrity were assessed with the
   NanoDrop ND-1000 UV-Vis Spectrophotometer (NanoDrop Technologies,
   Wilmington, DE, USA) and the Agilent 2100 Bioanalyzer (Agilent
   Technologies, Palo Alto, CA, USA). The 260/280 and 260/230 nm ratios
   were 2.15 ± 0.00 and 2.20 ± 0.05 in cells, respectively (n = 24, mean ±
   SEM). The RNA 6000 Nano LabChip kit (Agilent Technologies, Palo Alto,
   CA, USA) was used to evaluate the RNA integrity of the samples. The RNA
   integrity number (RIN) was 9.5 ± 0.1 (n = 24) in liver (mean ± SEM).

Direct RNA sequencing (RNA-seq)

   Direct RNA sequencing (RNA-seq) was used to screen for transcripts
   possibly affected by vitamin E (alpha tocopherol) and chlorpyrifos
   exposure in 24 Atlantic salmon hepatocyte cell cultures (about 7 mill.
   cells per culture). Poly (A) mRNA was isolated using magnetic beads
   with oligo (dT) from total RNA obtained from 24 cell cultures.
   Fragmentation buffer was added to shred mRNA to short reads. Using
   these short fragments (about 200 bp) as templates, random hexamer
   primers were applied to synthesize first-strand cDNA. Second-strand
   cDNA was synthesized using buffer, dNTPs, RNaseH, and DNA polymerase I.
   QiaQuick PCR extraction kit (Qiagen) was used to purify short
   double-stranded cDNA fragments. These fragments were then resolved with
   EB buffer for end reparation, added poly (A), and then ligated to the
   sequencing adapters. After agarose gel electrophoresis, the suitable
   fragments were selected for PCR amplification as templates. Finally,
   the libraries were sequenced using Illumina HiSeq 2000 (San Diego, CA,
   USA).

   An Atlantic salmon transcriptome assembled from liver tissue was used
   as a reference for alignment of the RNA-seq data (24 samples, 12
   millions reads per sample). Unigenes were annotated with Blastx
   alignment between unigenes and the databases of NR, NT, SwissProt,
   KEGG, COG and GO. The DESeq software package was used to screen for
   differentially expressed genes (DEGs). The DESeq package is based on
   the negative binomial distribution, and provides a method to test for
   differential expression by use of a shrinkage estimator for the
   variance. We used P-adjustment ≤ 0.5 and the absolute value of log[2]
   ratio ≥ 1 as the threshold to judge the significance of gene expression
   difference. All RNA-seq work was performed by staff at the Beijing
   Genome Institute (BGI, Hong Kong).

Metabolite profiling

   Global biochemical profiles were determined in 24 Atlantic salmon
   hepatocyte cell cultures. Each sample contained about 7 million cells.
   Samples were extracted and prepared for analysis using Metabolon’s
   standard solvent extraction method. 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) mass
   analyzer. The sample extract was split into two aliquots, dried, and
   then reconstituted in acidic or basic LC-compatible solvents, each of
   which contained 11 or more injection standards at fixed concentrations.
   One aliquot was analyzed using acidic positive ion optimized conditions
   and the other using basic negative ion optimized conditions in two
   independent injections using separate dedicated columns. Extracts
   reconstituted in acidic conditions were gradient eluted using water and
   methanol both containing 0.1% Formic acid, while the basic extracts,
   which also used water/methanol, contained 6.5mM Ammonium Bicarbonate.
   The MS analysis alternated between MS and data-dependent MS^2 scans
   using dynamic exclusion. The samples destined for GC/MS analysis were
   re-dried under vacuum desiccation for a minimum of 24 hours prior to
   being derivatized under dried nitrogen using
   bistrimethyl-silyl-triflouroacetamide (BSTFA). The GC column was 5%
   phenyl and the temperature ramp was from 40° to 300°C in a 16 minute
   period. Samples were analyzed on a Thermo-Finnigan Trace DSQ
   fast-scanning single-quadrupole mass spectrometer using electron impact
   ionization. Accurate 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
   measurements could be made on the parent ion as well as fragments. The
   typical mass error was less than 5 ppm. Ions with less than two million
   counts require a greater amount of effort to characterize.
   Fragmentation spectra (MS/MS) were typically generated in data
   dependent manner, but if necessary, targeted MS/MS could be employed,
   such as in the case of lower level signals. Instrument variability was
   4% for internal standards and total process variability for endogenous
   metabolites was 12%. The bioinformatics system consisted of four major
   components, the Laboratory Information Management System (LIMS), the
   data extraction and peak-identification software, data processing tools
   for QC and compound identification, and a collection of information
   interpretation and visualization tools for use by data analysts.
   Identification of known chemical entities was based on comparison to
   metabolomic library entries of purified standards. The metabolomics
   work was done by employees at Metabolon, USA.

Statistics

   Gene expression levels were calculated by using the Reads per kb per
   million reads (RPKM) method [[54]26]. Screening of differentially
   expressed genes (DEGs) was done with pairwise comparison using the
   DESeq software [[55]27]. GO annotation was obtained from blastx
   searches against the NR database of NCBI by using BLAST2GO with default
   parameters (blastx E10^−5). Pathways enrichment analysis was done by
   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. Analysis by two-way ANOVA
   identified biochemicals exhibiting significant interactions and main
   effects for experimental parameters of dietary nutrient or contaminant.
   Welch’s two-sample t-tests were used to identify
   biochemicals/metabolites that differed significantly between
   experimental groups (P<0.05). Correction for multiple testing was done
   with false discovery rate (FDR) using q-values (P-adj) [[56]28].
   Statistical analyses of the log-transformed data were performed with
   the program “R” [[57]29]. The functional pathway analyses were
   generated through the use of IPA [[58]30].

Results

Exposure experiment and cytotoxicity

   The different cell suspensions used in this study had cell viability
   between 89–96% as determined by the Trypan Blue exclusion method. After
   48 hours of exposure, no significant effect of chlorpyrifos treatment
   was observed on cytotoxicity measured with the xCELLigence system.
   Neither did vitamin E supplementation, alone nor in combination with
   chlorpyrifos, have any effect on cytotoxicity (data not shown).

Transcriptomics results

   On average, 12,085,418±62,567 single-end Illumina reads were sequenced
   from the cell culture samples (n = 24, mean±SEM). The reads were mapped
   to a liver Atlantic salmon transcriptome assembled from 65,863,720
   paired-end Illumina reads. Total mapped reads were 10,163,717±62,277 (n
   = 24, mean±SEM) or about 84%. The reference transcriptome was made from
   a pool of total RNA obtained from liver of four juvenile Atlantic
   salmon. Assembly of the transcriptome created 156,417 contigs with an
   average length of 266 nucleotides (nts) and 65,519 unigenes with an
   average length of 571 nts, as well as 45,855 singletons. The number of
   unigenes annotated in various databases were: NR: 32,699, NT: 41,578,
   Swiss-Prot: 29,033, KEGG: 23,356, COG: 9,486, and GO: 22,371.

   [59]Table 2 shows an overview of the number of differentially expressed
   genes (DEGs) according to the transcriptional profiling. The DESeq
   method was used with a log2 ratio ≥1 (>2-fold change) and FDR-adjusted
   P-values (P-adj) <0.50 and <0.10. In the pairwise comparison (A versus
   B), the former one (A) is considered as the control, and the latter one
   (B) is considered as the treatment. Exposure to 100 μM chlorpyrifos
   significantly upregulated 942 transcripts and downregulated 901
   transcripts. Treatment with 100 μM alpha tocopherol alone did not
   significantly affect any transcripts in the cells. Compared to the
   control, by adding alpha tocopherol to chlorpyrifos the number of
   significantly upregulated transcripts was lowered from 942 to 626 and
   the number of significantly downregulated transcripts was lowered from
   901 to 742. A direct comparison between the chlorpyrifos and vitamin E
   treatment groups showed 636 upregulated and 834 downregulated
   transcripts. The comparison between the vitamin E and vitamin E plus
   chlorpyrifos treatment groups produced 495 upregulated and 399
   downregulated transcripts. [60]S1 Dataset shows the gene lists from
   these pairwise comparisons. Using a stricter P-adj cut-off of <0.1, the
   corresponding numbers of significantly regulated genes from the
   pairwise comparisons were Control vs. CPF: 351 genes upregulated/182
   genes downregulated, Control vs. CPF+VitE: 228 upregulated/140
   downregulated, and CPF vs. VitE: 107 upregulated/306 downregulated.
   Since the primary goal of the transcriptional profiling was to identify
   biological processes, pathways and ontology categories, and not to
   identify specific biomarkers, gene lists created with the less strict
   P-adj cut-off was selected for downstream IPA analyses. With this
   approach, about 45% of the DEGs listed in [61]S1 Dataset and used for
   downstream analyses were identified with annotation (blastx search
   against the NR database, cut-off E10^–5).

Table 2. Number of affected transcripts and metabolites using two different
statistical stringencies.

   Statistical comparisons
   Significantly altered transcripts/metabolites Total transcripts Log2≥1
   P-adj<0.50 Transcripts (⇵) Total transcripts Log2≥1 P-adj<0.10
   Transcripts (⇵) Total metabolites P≤0.05 Metabolites (⇵) Total
   metabolites P<0.05<P-adj<0.10 Metabolites (⇵)
   Vit E 0 0|0 0 0|0 1 0|1 3 0|3
   Control
   CPF 1843 942|901 533 351|182 155 130|25 25 21|4
   Control
   Vit E + CPF 1368 626|742 368 228|140 117 81|36 36 18|18
   Control
   CPF 1470 636|834 413 107|306 154 135|19 30 23|7
   Vit E
   Vit E + CPF 894 495|399 221 164|57 87 58|29 51 37|14
   Vit E
   Vit E + CPF 0 0|0 0 0|0 32 0|32 28 1|27
   CPF
   [62]Open in a new tab

   CPF = Chlorpyrifos. Vit E = Vitamin E.

   KEGG pathway enrichment analysis identified JAK-STAT signaling (pathway
   ID ko04630) as the most significantly affected system by chlorpyrifos
   exposure in Atlantic salmon hepatocytes ([63]Fig. 1A), followed by
   Protein processing in endoplasmic reticulum (ko04141) ([64]Fig. 1B),
   Pathways in cancer (ko05200) and Osteoclast differentiation (ko04380).
   Statistically significant KEGG pathways (P<0.05) from gene set
   enrichment analysis are shown in [65]S2 Dataset. Pathways linked to
   typical human diseases and of little relevance for fish were omitted
   from the lists. Addition of vitamin E to chlorpyrifos appears to have a
   limited impact on these molecular mechanisms and others according to
   the KEGG pathway enrichment analysis. Antigen processing and
   presentation (ko04612) became more significantly affected by the
   inclusion of vitamin E in the exposure media, but most of the pathways
   appear on both lists although with slightly different significance
   levels. Vitamin E supplementation amended the following significantly
   changed pathways, Carbohydrate digestion and absorption (ko04973),
   Aldosterone-regulated sodium reabsorption (ko04960), Glycosaminoglycan
   biosynthesis—chondroitin sulfate (ko00532) (significantly affected by
   chlorpyrifos alone, but not with co-treatment with vitamin E), and
   Focal adhesion (ko04510), MAPK signaling pathway (ko04010), Ubiquitin
   mediated proteolysis (ko04120), Other types of O-glycan biosynthesis
   (ko00514), D-Glutamine and D-glutamate metabolism (ko00471), Regulation
   of actin cytoskeleton (ko04810) and ErbB signaling pathway (ko04012)
   (significantly affected with vitamin E supplementation, but not by
   chlorpyrifos alone).

Fig 1. Predicted cellular outcome of chlorpyrifos exposure (100 μM) in
Atlantic salmon hepatocytes based on affected transcripts.

   Fig 1
   [66]Open in a new tab

   Effects on A) Jak-STAT signaling (KEGG pathway ID ko04630) and B)
   Protein processing in endoplasmic reticulum (ko04141). Colored enzymes:
   green for downregulated genes, red for upregulated genes.

Metabolomics results

   A total of 28 samples were analyzed for metabolites, 24 cell culture
   samples and four exposure medium samples. Cell extracts (n = 6 per
   group) were normalized to Bradford protein values prior to analysis.
   From analysis of the dataset a total of 329 named metabolites were
   detected from the metabolite profiling. Vitamin E and chlorpyrifos
   interacted significantly on 15 metabolites ([67]Table 3), while 9
   metabolites were significantly affected by vitamin E alone and 196 by
   chlorpyrifos alone (2-way ANOVA). [68]Table 2 shows an overview of the
   number of differentially regulated metabolites between the four
   treatment groups determined by Welch's two-sample t-test and using two
   levels of statistical stringency (P<0.05 or P<0.05 and P-adj<0.1). With
   the former stringency, 130 metabolites were upregulated and 25
   downregulated by chlorpyrifos exposure compared to the control.
   Supplementation of vitamin E reduced the number of significantly
   upregulated metabolites 130 to 81, while at the same time increased the
   number of downregulated metabolites from 25 to 36. Compared to the
   control, vitamin E supplementation alone only significantly
   downregulated 1 metabolite (2-stearoylglycerophosphoinositol).
   Principal component analysis (PCA) revealed a separation between the
   control and chlorpyrifos-treated cells regardless of vitamin E
   treatment. Heterogeneity was observed within sample groups suggesting
   differences in basal metabolism between these primary-derived cells
   obtained from six different male fish. Chlorpyrifos-treated cells also
   sorted separately by hierarchical clustering compared to the control.
   This trend was independent of vitamin E suggesting this antioxidant may
   have a limited impact on the metabolic profile of these hepatocytes.
   Chlorpyrifos was detected in cell extracts ([69]Fig. 2A), but not
   exposure media suggesting this metabolite was below the limit of
   detection in these samples. Alpha tocopherol was detected in exposure
   media ([70]Fig. 2B) and in the cells ([71]Fig. 2C). Pathway-specific
   heat maps of affected metabolites, with fold changes and significance
   levels, are shown in [72]S3 Dataset. Chlorpyrifos exposure induced
   changes in metabolites linked to four major systems in the hepatocytes.
   These were energy metabolism, lipid metabolism, BCAA/glutathione
   metabolism and nucleotide metabolism.

Table 3. Biochemicals with interaction effects between chlorpyrifos (CPF) and
vitamin E (Vit E).

   Two-Way ANOVA
   Super Pathway Sub Pathway Biochemical Name Platform Comp ID KEGG HMDB
   CPF Main Effect P-value Vit E Main Effect CPF:Vit E Interaction P-value
   Amino Acid Phenylalanine and Tyrosine Metabolism homogentisate LC/MS
   neg 521 [73]C00544 HMDB00130 0.0083 0.0370
   Methionine, Cysteine, SAM and Taurine Metabolism homocysteine GC/MS
   40266 [74]C00155 HMDB00742 0.8590 0.0143
   Peptide Dipeptide alanylarginine LC/MS neg 37096 0.0058 0.0291
   alanylisoleucine LC/MS pos 37118 0.0038 0.0229
   alanylleucine LC/MS pos 37093 0.0021 0.0347
   alanylphenylalanine LC/MS pos 38679 0.0000 0.0334
   alanyltyrosine LC/MS pos 37098 0.0000 0.0255
   alpha-glutamyltyrosine LC/MS pos 40033 0.0004 0.0304
   leucylalanine LC/MS pos 40010 0.0002 0.0462
   leucylarginine LC/MS neg 40028 0.0000 0.0174
   methionylmethionine LC/MS neg 40696 0.0031 0.0117
   seryltyrosine LC/MS pos 42077 0.0031 0.0211
   valyllysine LC/MS neg 41384 0.0000 0.0333
   valyltyrosine LC/MS neg 40697 0.0002 0.0366
   Lipid Fatty Acid, Monohydroxy 13-HODE + 9-HODE LC/MS neg 37752 0.6442
   0.0093
   [75]Open in a new tab

Fig 2. Scaled intensity levels of A) chlorpyrifos and B) alpha tocopherol
(vitamin E) in Atlantic salmon hepatocytes exposed to chlorpyrifos (100 μM),
alpha tocopherol (100 μM) or combined (100 μM of both).

   [76]Fig 2
   [77]Open in a new tab

   The figures show boxplots with median, upper and lower quartiles, max
   and min distribution, mean value (+) and extreme data points (ο).

   Chlorpyrifos exposure increased the accumulation of multiple free amino
   acids in the hepatocytes. Elevated levels of valine, isoleucine, and
   leucine in chlorpyrifos-treated cells were accompanied by an
   accumulation of the related alpha-keto acids 4-methyl-2-oxopentanoate
   and 3-methyl-2-oxovalerate ([78]Fig. 3A). Vitamin E supplementation
   partially restored dipeptide levels in chlorpyrifos treated cells.
   Chlorpyrifos exposure depleted the levels of glutathione (GSH) and
   oxidized glutathione (GSSG), and increased the levels of cystine
   (formed by the oxidation of two cysteine molecules that covalently link
   via a disulfide bond) ([79]Fig. 3A). In addition, increased levels of
   cystathionine, homocysteine and 5-oxoproline and diminished levels of
   2-aminobutyrate and ophthalmate were seen in chlorpyrifos-exposed
   cells.

Fig 3. Predicted cellular outcome of chlorpyrifos exposure (100 μM) in
Atlantic salmon hepatocytes based on affected metabolites.

   [80]Fig 3
   [81]Open in a new tab

   Effects on A) BCAA/glutathione metabolism B) lipid metabolism, C)
   energy metabolism and D) nucleotide metabolism. The figures show
   boxplots with median, upper and lower quartiles, max and min
   distribution, mean value (+) and extreme data points (ο) of scaled
   intensities. Values are normalized by Bradford protein concentration.

   The long chain fatty acids palmitate and stearate accumulated in
   chlorpyrifos-treated hepatocytes compared to the controls ([82]Fig.
   3B). Vitamin E treatment was able to partially restore fatty acid
   levels to near control levels. The accumulation of these metabolites in
   chlorpyrifos-treated cells may reflect a change in the hydrolysis of
   complex lipids such as triglycerides as supported by the accumulation
   of monoacylglycerols such as 1-linoleoylglycerol and
   1-arachidonylglycerol. The accumulation of these metabolites was
   modestly impacted by vitamin E supplementation. The polyunsaturated
   fatty acids docosapentaenoate, docosahexaenoate, and arachidonate were
   diminished in the presence of chlorpyrifos. Decreased levels of these
   metabolites indicate a conversion to lipid peroxidation products such
   as 13-HODE and 9-HODE that can be indicative of increased oxidative
   stress and are PPAR ligands associated with inflammatory hyperalgesia.
   Vitamin E supplementation also prevented chlorpyrifos-induced
   cholesterol accumulation in the cells.

   For energy metabolism, chlorpyrifos exposed cells exhibited reduced
   levels of glycogen metabolites maltopentaose, maltotriose and maltose
   compared to controls ([83]Fig. 3C). Vitamin E supplementation in the
   presence of chlorpyrifos was unable to restore glycogen metabolite
   levels. Alterations in glycogen may restrict glucose availability as
   evidenced by reduced levels of glucose in these cells. The glycolytic
   metabolites glucose 6-phosphate (G6P), fructose 1,6-bisphosphate,
   3-phosphoglycerate, and lactate were reduced in the presence of the
   pesticide, regardless of vitamin E treatment. The TCA cycle
   intermediates citrate, alpha-ketoglutarate, and succinate significantly
   accumulated in the presence of chlorpyrifos.

   Following chlorpyrifos exposure, the purine degradation products
   inosine, hypoxanthine, adenine, and xanthine accumulated in hepatocytes
   ([84]Fig. 3D). Elevated levels of these metabolites can contribute to
   altered redox homeostasis since the generation of xanthine is
   accompanied by the production of hydrogen peroxide (H[2]O[2]). Purine
   catabolism may also be a component of the homeostatic response of the
   mitochondria to oxidative stress. Alternatively, these findings reflect
   increased nucleic acid availability potentially resulting from
   decreased cell growth and/or limited energy generation and excess AMP
   levels. Elevated levels of the pyrimidines cytidine and thymidine in
   chlorpyrifos treated cells were accompanied by an accumulation of
   uracil that may be indicative of an increase in nucleotide catabolism.
   These metabolic imbalances were not alleviated by vitamin E
   supplementation.

Ingenuity pathway analysis (IPA)

   IPA Core analysis and the IPA Compare function were used for evaluation
   of biological processes, pathways and networks. In order to use IPA,
   all identifiers must be recognized as mammalian homologs. Some
   fish-specific genes obviously cannot be given human ortholog names
   recognized by IPA, and thus cannot be included in IPA-Core analysis. To
   search for ameliorating effects of vitamin E on chlorpyrifos toxicity,
   the pairwise “control versus chlorpyrifos” and “control versus
   chlorpyrifos and vitamin E” comparisons were more closely examined. The
   number of identifiers included in the “control versus chlorpyrifos” and
   “control versus chlorpyrifos and vitamin E” IPA Core analyses were 640,
   77 and 716, and 482, 64 and 565, respectively (transcripts,
   metabolites/biochemicals and combined). Heat maps generated with the
   IPA Compare function based on activation score from the transcriptomics
   analyses suggested that vitamin E supplementation ameliorated effects
   induced by chlorpyrifos related to cell survival and viability, bone
   cell formation-linked mechanisms, fatty acids (uptake of palmitic acid,
   concentration of acylglycerol) and amino acids (transport of L-amino
   acid), as well as numerous pathways linked to cancer and other
   mechanisms mainly of relevance for human health ([85]S1 Table). A
   similar analysis by using the IPA Core analyses with metabolite
   identifiers suggested that vitamin E supplementation lowered the impact
   on free radical generation and detoxification (synthesis of nitric
   oxide, concentration of glutathione) and corresponding effects as
   suggested by the transcriptional data on fatty acid metabolism and
   amino acid transport. By combining the significantly affected
   transcripts and metabolites, in addition to effects on cell survival
   and viability, predicted activation of functions linked to fatty acids
   (accumulation of lipids ([86]Fig. 4A), uptake of fatty acid, uptake of
   palmitic acid, accumulation of triacylglycerol) and predicted
   inhibition of functions linked to carbohydrates (metabolism of
   polysaccharide ([87]Fig. 4B), transport and oxidation of
   monosaccharide, oxidation of D-glucose) were the most distinct effect
   of vitamin E supplementation.

Fig 4. Ameliorating effects of vitamin E (100 μM) on chlorpyrifos (100 μM)
toxicity in Atlantic salmon hepatocytes according to predicted activation and
inhibition of biological functions.

   [88]Fig 4
   [89]Open in a new tab

   Based on combined lists of transcripts and metabolites. Data from IPA
   Core and Compare analyses. The legends show activation and inhibition
   states, and contribution of up-regulated and down-regulated molecules.
   A) Accumulation of lipid and B) Metabolism of polysaccharide.

Discussion

   This work shows that alpha tocopherol (vitamin E) to a modest degree
   can affect chlorpyrifos toxicity in Atlantic salmon liver cells. The
   cellular response as predicted by the metabolite outcome suggests that
   the main effects of chlorpyrifos exposure were on energy metabolism,
   lipid metabolism, BCAA/glutathione metabolism and nucleotide
   metabolism. Jak-STAT signaling was the most strongly affected pathway
   by chlorpyrifos exposure according to the transcriptional profiling.
   Vitamin E supplementation could only to a certain degree rescue the
   cells from the negative effects of the organophosphate pesticide.

   In the nervous system, one of the main targets of chlorpyrifos toxicity
   in animals [[90]10,[91]31], the negative effects of the pesticide can
   lead to development of oxidative stress [[92]32]. Numerous studies both
   with mammalian and fish models suggest that oxidative stress is one of
   the main outcomes of chlorpyrifos exposure [[93]15]. As predicted,
   oxidative stress mechanisms were also affected by chlorpyrifos exposure
   in the Atlantic salmon hepatocytes. Depleted glutathione (GSH) and
   oxidized glutathione (GSSG) in the presence of chlorpyrifos indicate a
   decline in total glutathione availability, while elevated levels of
   cystine (formed by the oxidation of two cysteine molecules that
   covalently link via a disulfide bond) suggests free radical exposure.
   Alternatively, elevated cystathionine and homocysteine levels coupled
   with diminished 2-aminobutyrate and ophthalmate levels suggest a
   deficit in cysteine biogenesis and consequently the capacity to
   replenish glutathione. In support, higher levels of 5-oxoproline
   following pesticide treatment is indicative of the import and
   degradation of gamma-glutamyl amino acids to replenish GSH. Vitamin E
   administration was unable to restore cysteine metabolism in this study.
   Elevated accumulation of the purine degradation products inosine,
   hypoxanthine, adenine, and xanthine following chlorpyrifos exposure
   indicate altered redox homeostasis since the generation of xanthine is
   accompanied by the production of hydrogen peroxide (H[2]O[2]) [[94]33].
   Published studies also demonstrate that purine catabolism may be a
   component of the homeostatic response of the mitochondria to oxidative
   stress [[95]34].

   At the transcriptional level, chlorpyrifos exposure significantly
   affected the Gene Ontology “Response to oxidative stress” (GO:0006979)
   genes antioxidant 1 copper chaperone (ATOX1), protein phosphatase 1
   regulatory subunit 15B (PPP1R15B) and nuclear factor erythroid
   2-related factor 2 (NFE2L2). The latter gene, NFE2L2, is a
   transcription activator that binds to antioxidant response elements
   (ARE) in the promoter regions of target genes important for the
   coordinated regulation of genes in response to oxidative stress
   [[96]35]. NFE2L2, which was downregulated by chlorpyrifos exposure, did
   not show altered transcription in cells co-treated with vitamin E. In
   addition, the B-cell translocation gene 1, anti-proliferative (BTG1)
   and NF-kappa-B 1 p105 subunit (NFKB1) genes were significantly affected
   by chlorpyrifos and vitamin A co-treatment but not by chlorpyrifos
   alone, while vitamin E supplementation counteracted the
   chlorpyrifos-induced ATOX1 response. The NFKB1 encodes a transcription
   factor present in almost all cell types involved in multiple cellular
   responses to stimuli such as cytokines and stress [[97]35]. Its
   downregulation in cells co-treated with chlorpyrifos and vitamin E but
   not in pesticide-treated cells alone may indicate ameliorating effects
   of the antioxidant on mechanisms linked to apoptosis or
   proteasome-dependent degradation of proteins. Accordingly, the ATOX1
   protein functions as an antioxidant against superoxide and hydrogen
   peroxide [[98]35], and the differential transcription in cells given
   vitamin E may suggest a protective effect of the antioxidant.
   Chlorpyrifos exposure also induced pyruvate dehydrogenase kinase
   isozyme 2 (PDK2), a serine/threonine kinase functionally belonging in
   the Gene Ontology “Cellular response to reactive oxygen species”
   (GO:0034614) that plays a role in the regulation of cell proliferation
   and in resistance to apoptosis under oxidative stress [[99]35]. No
   effect of vitamin E supplementation was however seen on PDK2
   transcription in the cells. CuZn superoxide dismutase (SOD1) and
   catalase (CAT), suggested markers for longer-term chlorpyrifos exposure
   (15 days) in the Nile tilapia (Oreochromis niloticus) by Oruc
   [[100]36], were not among the affected transcripts in cultured Atlantic
   salmon hepatocytes. To see effects on typical oxidative stress response
   genes, fish studies suggest that longer-term chlorpyrifos exposure is
   needed [[101]37]. Overall, the transcriptional data including
   significantly affected pathways ([102]S2 Dataset) suggest that the
   antioxidant had only a modest effect on chlorpyrifos-induced oxidative
   damage leading to apoptosis.

   One of the most pronounced effects of chlorpyrifos exposure on
   metabolites was the accumulation of multiple free amino acids in the
   hepatocytes. The accumulation of multiple free amino acids in
   chlorpyrifos treated cells may reflect decreased utilization for
   protein synthesis or an increase in proteolysis considering multiple
   dipeptides significantly accumulated in these cells. Elevated levels of
   valine, isoleucine, and leucine in chlorpyrifos-treated cells were
   accompanied by an accumulation of the related alpha-keto acids
   4-methyl-2-oxopentanoate and 3-methyl-2-oxovalerate that may also be
   indicative of degradation. These findings may reflect oxidative damage
   of proteins and subsequent degradation. Vitamin E supplementation
   partially restored dipeptide levels in chlorpyrifos treated cells and
   this was the strongest effect of vitamin E observed in the current
   study. Significant interaction effects between chlorpyrifos and vitamin
   E were observed for 12 dipeptides. Collectively, these findings are in
   agreement with published studies demonstrating chlorpyrifos induces
   oxidative stress in multiple model systems
   [[103]36,[104]37,[105]38,[106]39,[107]40]. Alpha tocopherol
   supplementation at the studied concentration did not counteract these
   effects, suggesting that vitamin E, except its ability to partially
   restore dipeptide levels, has a relatively limited ability to rescue
   chlorpyrifos-induced oxidative stress in salmon hepatocytes.

   Chlorpyrifos exposure affected the levels of several long chain fatty
   acids. The accumulation of these metabolites in chlorpyrifos treated
   cells indicates a change in the hydrolysis of complex lipids such as
   triglycerides as supported by the accumulation of monoacylglycerols
   such as 2-linoleoylglycerol and 1-arachidonylglycerol. Vitamin E
   treatment was able to partially restore the levels of these fatty acids
   to near control levels. Altered levels of lysolipids indicate a
   difference in membrane remodeling in the presence of pesticide. The
   accumulation of these metabolites was only modestly impacted by vitamin
   E supplementation suggesting a partial restoration of lipid
   homeostasis. Elevated lipid levels indicate increased eta-oxidation as
   suggested by the accumulation of acetylcarnitine (contributes to the
   movement of acetyl CoA in the mitochondria during lipid oxidation), in
   line with earlier mammalian studies [[108]39]. In contrast to long
   chain fatty acids, the polyunsaturated fatty acids docosapentaenoate,
   docosahexaenoate, and arachidonate were diminished in the presence of
   chlorpyrifos. Lower polyunsaturated fatty acids indicate utilization of
   these precursors for the generation of pro- and anti-inflammatory
   eicosanoids. Furthermore, decreased levels of these metabolites may be
   indicative of conversion to lipid peroxidation products such as 13-HODE
   and 9-HODE that can be indicative of increased oxidative stress and are
   peroxisome proliferator-activated receptor (PPAR) ligands associated
   with inflammatory hyperalgesia. 13-HODE and 9-HODE are oxidative
   metabolites of the essential fatty acid linoleic acid [[109]35] and
   these were the only lipid metabolism biochemicals that showed a
   significant interaction effect for chlorpyrifos and vitamin E. In a
   previous study in which Atlantic salmon hepatocytes were exposed to
   chlorpyrifos [[110]22], several putatively annotated eicosanoids in
   different lipid metabolism pathways were affected by the exposure. We
   were however not able to reproduce the effect on these metabolites in
   the current experiment, although the suggested overall effect on lipid
   metabolism was the same. A predicted inhibition of release of
   eicosanoids, due to altered levels of arachidonic acid, D-glucose,
   D-mannose, glycine and cholesterol after chlorpyrifos exposure, was
   enhanced by vitamin E supplementation in the current study. Compared to
   the cells exposed to chlorpyrifos alone, vitamin E supplementation
   contributed to a predicted inhibition of eicosanoid synthesis.

   It is well known that many lipid-soluble contaminants, including
   organophosphorous pesticides, can contribute to accumulation of lipids
   in fish liver [[111]41]. In crucian carp (Carassius auratus gibelio),
   for example, Xu et al. [[112]42] found increased levels of
   triglycerides in the liver after exposure to the organophosphorous
   pesticide trichlorfon. Recent studies have shown a strong correlation
   between chemical exposure and steatosis (fatty liver) [[113]43]. We
   have earlier shown that the organochlorine insecticide endosulfan
   induces steatosis in Atlantic salmon hepatocytes [[114]44], while Sun
   et al. [[115]45] showed a similar response in male rockfish
   (Sebastiscus marmoratus) exposed to the triazole-containing fungicide
   paclobutrazol. As shown in [116]Fig. 3A, a number of genes who’s
   proteins are linked to increased accumulation of lipids displayed
   differential expression in this study. Of these, it is interesting to
   note that while vitamin E prevented the chlorpyrifos-induced increased
   expression of PPARA, no effect was seen on the PPAR beta 1A (PPARB1A,
   an ortholog to mammalian PPARD).

   Impaired metabolism of a number of drugs has been linked to steatosis,
   suggesting an association between increased lipid deposition and
   impaired cytochrome P450 (CYP) enzymes [[117]46]. In carp (Cyprinus
   carpio), it has been shown that long-term exposure to chlorpyrifos
   results in CYP1A mRNA induction and increased EROD activity in the
   liver [[118]47]. Surprisingly, the highly inducible CYP1A gene was not
   among the gene transcripts significantly induced by chlorpyrifos
   exposure in the exposed liver cells. According to the transcriptional
   data, only two gene transcripts, both annotated to cytochrome P450 2K1,
   were differentially affected by chlorpyrifos exposure. Co-treatment
   with vitamin E prevented this effect.

   One possible explanation on how vitamin E might help prevent
   accumulation of lipids in fish liver is disturbance of cholesterol
   transport and uptake, as no cholesterol accumulation was observed in
   pesticide-treated cells. The transcriptional data showed that
   chlorpyrifos-induced downregulation of low-density lipoprotein receptor
   (LDLR) was prevented by vitamin E treatment. LDLR binds low-density
   lipoprotein (LDR), the major cholesterol-carrying lipoprotein of
   plasma, and transports it into cells by endocytosis [[119]35].
   Similarly, hormone-sensitive lipase (LIPE), which primarily hydrolyzes
   stored triglycerides to free fatty acids [[120]35], and fatty
   acid-binding protein 3 (FABP3) which is involved in uptake,
   intracellular metabolism and transport of long-chain fatty acids
   [[121]35], were upregulated by chlorpyrifos treatment but not in cells
   co-treated with vitamin E. The antioxidant appears to have less effect
   on membrane-associated gene proteins such as sphingosine 1-phosphate
   receptor 2 (S1PR2) and phospholipase A2, Group XIIB (PLA2G12B) that
   were affected by chlorpyrifos treatment.

   One of the main findings in the current study was that chlorpyrifos
   exposure affected energy metabolism in the cells. Chlorpyrifos exposed
   cells exhibited reduced levels of the several glycogen metabolites.
   Vitamin E supplementation in the presence of chlorpyrifos was unable to
   restore glycogen metabolite levels. Alterations in glycogen may
   restrict glucose availability as evidenced by reduced levels of glucose
   in the cells. Consequently, the glycolytic metabolites glucose
   6-phosphate (G6P), fructose 1,6-bisphosphate, 3-phosphoglycerate, and
   lactate were reduced in the presence of the pesticide. Alternatively,
   lower glucose levels may suggest a difference in transport as supported
   by rat studies showing that chlorpyrifos inhibits glucose uptake in
   liver [[122]38]. Consequently, the glycolytic metabolites glucose
   6-phosphate (G6P), fructose 1,6-bisphosphate, 3-phosphoglycerate, and
   lactate were reduced in the presence of the pesticide, regardless of
   vitamin E treatment. Chronic exposure to organophosphate pesticides can
   lead to deleterious effects on carbohydrate metabolism.
   Carbohydrate-metabolizing organs, such as the liver, can be affected by
   organophosphate pesticides through altered glycolysis, gluconeogenesis,
   glycogenesis, and glyconeogenesis [[123]48]. [124]Fig. 3B summarizes
   which gene transcripts and metabolites that were involved in
   polysaccharide metabolism as identified by the IPA analyses. In
   contrast to glycolytic intermediates, the TCA cycle intermediates
   citrate, alpha-ketoglutarate, and succinate significantly accumulated
   in the presence of chlorpyrifos regardless of vitamin E
   supplementation. This imbalance in TCA metabolites may be indicative of
   mitochondrial dysfunction and potentially reflect oxidative inhibition
   of mitochondrial enzymes such as aconitase. Similarly, the reduced
   activity of liver mitochondrial isocitrate dehydrogenase (IDH) has been
   correlated with organophosphate toxicity in other models
   [[125]49,[126]50]. Taken together, these findings suggest that
   chlorpyrifos may limit the energetic capacity of salmon hepatocytes,
   and that vitamin E supplementation has limited ability to restore these
   mechanisms.

   Interestingly, the IPA evaluation suggested that several pathways
   normally linked to the central nervous system were affected in liver by
   chlorpyrifos exposure. This finding is not surprising given that one of
   the main effects of chlorpyrifos exposure in animals is inhibition of
   AChE activity [[127]31]. Two major cholinesterase enzymes are found in
   fish. AChE is most common in brain and muscle, while
   butyrylcholinesterase (BChE) is found mainly in plasma and in the liver
   [[128]51]. Relatively high AChE activity has been reported in fish
   liver [[129]52]. A significant reduction in hepatic AChE activity and
   mRNA level has been reported in fish exposed to chlorpyrifos [[130]53].
   The significant effect seen on Jak-STAT signaling by chlorpyrifos
   exposure may be due to damage inflicted by the bioactivated oxon on
   liver cholinesterases or reflect alterations in protein phosphorylation
   pathways [[131]54]. A similar response has been reported in brain of
   mice exposed to chlorpyrifos [[132]55].

   Chlorpyrifos is a known endocrine disruptor. Two transcripts annotated
   to genes associated with the Gene Ontology “Response to estrogen
   stimulus” (GO:0043627) were significantly affected by pesticide
   treatment. Both the TGF beta receptor-1 (TGFBR1) and the
   Cbp/p300-interacting transactivator 2 (CITED2) were downregulated by
   chlorpyrifos exposure. Vitamin E supplementation did however not alter
   this response. In addition, two gene transcripts associated with the
   Gene Ontology “Estrogen receptor binding” (GO:0030331) were
   differentially regulated. ATP-dependent RNA helicase DDX5 (DDX5) was
   upregulated by chlorpyrifos exposure, while nuclear
   receptor-interacting protein 1 (NRIP1) was downregulated by the
   pesticide. The latter response was ameliorated by vitamin E treatment,
   however, the TGF-beta signaling pathway response was not modified by
   the antioxidant. These findings suggest that chlorpyrifos may affect
   mechanisms linked to endocrine disruption in Atlantic salmon liver
   cells.

   The transcriptional and metabolite profiling approaches give
   complementary outcomes, supported by the IPA analyses. According to the
   transcriptional data, Jak-STAT signaling, Protein processing in
   endoplasmic reticulum and Pathways in cancer were the three most
   significantly affected KEGG pathways by chlorpyrifos exposure. These
   results suggest that the pesticide induced cellular damage leading to
   apoptosis. Significant effects were seen on several DEGs belonging to
   the Gene Ontology “Induction of apoptosis” (GO:0006917). These included
   genes such as GTP cyclohydrolase 1 (GCH1), serine/threonine-protein
   kinase 24 (STK24), Krueppel-like factor 10 (KLF10), rho
   GTPase-activating protein 7 (DLC1), TGFBR1, etoposide-induced protein
   2,4 homolog (EI24), mitochondrial ubiquitin ligase activator of NF-kB
   (MUL1), as well as apoptosis regulator BCLX. This result fits nicely
   with the observed accumulation of free amino acid as shown by the
   metabolomics data. Branched amino acids, such as valine, isoleucine,
   and leucine discussed above, can induce apoptosis in animal cells
   [[133]56]. Jak-STAT signaling has also been linked to altered lipid
   metabolism [[134]57]. As described above, one of the main effects of
   chlorpyrifos exposure was induction of steatosis. Intracellular fat
   accumulation has been shown to induce endoplasmatic stress and disrupt
   Jak-STAT signaling in human cell cultures [[135]58], further
   demonstrating the complementary nature of the two methods. As a key
   regulator of numerous cytokines, growth factors and hormones, the
   JAK-STAT pathway [[136]57] can have a profound effect on energy
   metabolism in vivo [[137]57]. This study shows a similar outcome in
   cell cultures. Applied together, transcriptional and metabolite
   profiling therefore appear to be complementary by nature.

   In conclusion, this study shows that vitamin E supplementation may help
   protect cells against chlorpyrifos-induced toxicity by restoring
   dipeptide levels, preventing accumulation of lipids and modifying
   carbohydrate metabolism. Both the transcriptional and metabolite
   profiling suggest that this effect is relatively modest, and that
   vitamin E supplementation will not rescue the cells completely.
   Although chlorpyrifos exposure clearly induced oxidative stress in the
   hepatocytes, the protective effect of vitamin E was minor. The study
   also demonstrates the complementary nature of transcriptional and
   metabolite profiling, with the more detailed transcriptional responses
   supplementing the cellular outcome as predicted by the metabolite
   changes.

Supporting Information

   S1 Table. Heat map generated with the IPA Compare function using both
   the transcriptomics and metabolomics data.

   (XLSX)
   [138]Click here for additional data file.^ (42.2KB, xlsx)
   S1 Dataset. Pairwise comparisons of transcriptional data.

   (XLSX)
   [139]Click here for additional data file.^ (645.2KB, xlsx)
   S2 Dataset. Statistically significant KEGG pathways from gene set
   enrichment analysis.

   (XLSX)
   [140]Click here for additional data file.^ (54.3KB, xlsx)
   S3 Dataset. Pathway-specific heat maps of affected metabolites.

   (XLSX)
   [141]Click here for additional data file.^ (128.5KB, xlsx)

Acknowledgments