Perfluorooctane Sulfonamide-Mediated Modulation of Hepatocellular

May 17, 2012 - Daniela Metro , Roberta Tardugno , Mattia Papa , Carlo Bisignano , Luigi ... Augustine Arukwe , Maria V. Cangialosi , Robert J. Letcher...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/crt

Perfluorooctane Sulfonamide-Mediated Modulation of Hepatocellular Lipid Homeostasis and Oxidative Stress Responses in Atlantic Salmon Hepatocytes Ane Marit Wågbø,† Maria V. Cangialosi,‡ Nicola Cicero,‡ Robert J. Letcher,§ and Augustine Arukwe*,† †

Department of Biology, Norwegian University of Science and Technology (NTNU), Høgskoleringen 5, 7491 Trondheim, Norway Department of Food and Environmental Science “Prof. G. Stagno d’Alcontres”, University of Messina, Salita Sperone 31, 98166, S. Agata, Messina, Italy § Ecotoxicology and Wildlife Health Division, Environment Canada, National Wildlife Research Centre, Carleton University, Ottawa, ON K1A 0H3, Canada ‡

ABSTRACT: We have investigated the effects of perfluorooctane sulfonamide (PFOSA) on cellular functions and lipid homeostasis (including β-oxidation) in salmon primary hepatocytes. Salmon hepatocytes were exposed to PFOSA at 0 (control), 2, 20, and 50 μM for 12 and 24 h. Fatty acids (FAs) and lipids were determined by GC-MS; FA elongase (FAE), Δ5-desaturase (FAD5), Δ6-desaturase (FAD6), peroxisome proliferator-activated receptors (PPARs), acyl coenzyme A (ACOX-1), glutathione peroxidase (GPx), catalase (CAT), and glutathione S-transferase (GST) mRNA were analyzed using qPCR. GST activity was analyzed by biochemical assays using 1-chloro-2,4-dinitrobenzene (CDNB) as substrate. Our data showed that PFOSA produced significant changes in FA composition that predominantly involved a decrease (at 12 h) and an increase (at 24 h) in FA methyl esters (FAMEs), MUFA, total PUFA, and (n-3 and n-6) PUFA. Particularly, an increase of α-linolenic acid (ALA; 18:3n-3), eicosapentaenoic acid [EPA; 20:5n-3], and arachidonic acid [ARA: 20:4n-6] with associated increase in FAE, FAD5, and FAD6 mRNA were observed after PFOSA exposure, while cis-13,16-docosadienoic acid (22:2) was significantly decreased. PFOSA produced apparent concentration-dependent increase of PPARα and PPARγ. CAT, GPx, and GST mRNA show that PFOSA produced concentration- and time-specific increase of CAT and GST, but no changes in GST activity were observed. In general, these responses indicate that PFOSA evokes deleterious effects on cellular lipid homeostasis and transcriptional responses that regulate cellular oxidative homeostasis in salmon hepatocytes. precursor (or “PreFOS”) production.4 Hydrolysis of POSF yields PFOS and its salts, whereas reaction with methyl or ethylamines yields the alkyl substituted sulfonamides: N-methyl perfluorooctane sulfonamide (NMeFOSA) and N-ethyl perfluorooctane sulfonamide (NEtFOSA), respectively. Dealkylation of NMeFOSA and NEtFOSA can generate perfluorooctane sulfonamide (PFOSA), which is the ultimate PFOS

1. INTRODUCTION Perfluorinated compounds (PFCs) such as perfluorinated sulfonic (PFSAs) and carboxylic acids (PFCAs), and/or their precursors are used in various industrial and consumer products such as fluorinated polymers, surfactants, insecticides, and aqueous fire-fighting foams, and have been manufactured for over 50 years.1 Starting in 1949, the 3M Company was the major producer of perfluorooctane sulfonyl fluoride (POSF), and its primary use was as the starting material for all subsequent perfluorooctane sulfonate (PFOS) and PFOS © 2012 American Chemical Society

Received: March 12, 2012 Published: May 17, 2012 1253

dx.doi.org/10.1021/tx300110u | Chem. Res. Toxicol. 2012, 25, 1253−1264

Chemical Research in Toxicology

Article

location of the first double bond in the third (n-3) or sixth (n6) position from the methyl end of the aliphatic carbon chain.19 Of the PUFAs, α-linolenic acid (ALA: 18:3n-3) can be desaturated and elongated to first form eicosapentaenoic acid (EPA: 20:5n-3) and thereafter to docosahexaenoic acid (DHA: 22:6n-3), whereas the n-6 FA, linolelaidic acid (18:2n-6), can be desaturated and elongated to arachidonic acid (ARA: 20:4n6) and thereafter to docosapentaenoic acid (DPA: 22:4n-6; see Figure 1 for a complete overview of this process). Peroxisomes

precursor. However, they voluntarily phased out their production in 2002.2 Regardless, PFOS levels in the biotic environment continue to be rather high and in some cases are not necessarily decreasing.1,3 This is likely due, in part, to continuing small scale production of ∼600 tonnes a year of C8 PFC chemistry in, for example, China and Brazil, but moreover due to large environment reservoirs of “PreFOS” such as PFOSA and N-alkylated-PFOSAs.2 PFOSA has received little attention over the years but is randomly distributed in certain species and locations and has been detected worldwide in fish, mammals, birds, and humans at concentrations in the range of 1−100 ng/g wet weight of liver tissue.3−5 Most of the toxicological effects studies of PFCs on bioaccumulative PFCAs and PFSAs, have been derived from mammalian systems, and these or similar effects are yet to be demonstrated in lower vertebrates, such as fish. For example, PFCAs and PFSAs and their precursors have been shown to exert a variety of biological effects, including lipid homeostasis and peroxisome proliferation, hepatomegaly, immunotoxicity, uncoupling of mitochondrial oxidative phosphorylation, developmental toxicity, reduction of thyroid hormone circulation, necrosis, and down-regulation of hepatic transporters and tumors.6−9 PFOSA appears to mediate greater toxicity and exhibit substantially different effects compared to other PFCs.10 In mammalian systems, there is evidence showing that PFOSA is metabolically degraded and can form PFOS metabolic products at a slow rate and that PFOSA can undergo enterohepatic circulation and is also a potent mediator of oxidative stress.10,11 It has also been suggested that the reduced hydrophilicity of PFOSA could facilitate passage across cell membranes, including the placenta and blood−brain barriers, to achieve high intracellular levels.10 The regulation of energy homeostasis is critical for normal physiology and survival, and disruption of this balance often leads to chronic disease state.12 Fatty acids (FAs) are energyrich and ubiquitous biological molecules that are used as metabolic fuels, covalent regulators of signaling molecules, and as essential components of cellular membranes.13 FAs act as signaling molecules involved in regulating the expression of genes that mostly encode proteins with roles in FA transport or metabolism.14 Thus, partitioning into oxidative versus synthetic pathways is critical for cell function and must be tightly regulated.12 Chemically mediated changes in the composition of lipids will affect many biological processes taking place in the body, including lipogenesis, lipid deposition and storage, lipid transport by lipoproteins, and FA uptake in tissues.15 The ability to regulate FA pools is essential for normal homeostasis,13 and peroxisome proliferator-activated receptors (PPARs) are known to be critical regulators of lipid homeostasis by controlling the balance between burning and storage of long FAs.16 PPARs are ligand-dependent transcription factors belonging to the nuclear hormone receptor superfamily.16 They (PPARs) exert pleiotropic responses by regulating energy homeostasis, adipose tissue differentiation and maintenance, cell proliferation, and tissue repair.17 PPAR activities are consequently changed in accordance with a wide variety of physiological conditions, mediated through the ubiquitin−proteasome degradation system and extracellular signaling pathways and kinases that lead to receptor phoshorylation.18 FAs in fish tissues are present in different lipid classes and with different functions. There are two classes of long chain polyunsaturated FAs (PUFAs), n-3s and n-6s, based on the

Figure 1. Simplified pathways for the synthesis and metabolism of essential fatty acids. Note that the omega-3 and -6 pathways compete for the same desaturase or elongase enzymes.

are organelles that produce and degrade hydrogen peroxide (H2O2) and are involved in FAs metabolism,20 and these biological functions are mediated through the PPARs. An increase in the production of H2O2 due to exaggerated FA βoxidation may produce harmful effects in the organism.21 Superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx) are enzymes that play significant roles as antioxidants, and their increased expression and activity are indicative of oxidative stress.22 Thus, optimal adaptation during limited oxygen requires the induction of antioxidant and associated enzymes to reduce potential damage during oxygen reintroduction, such as lipid peroxidation.23 Previous studies on the effects of PFCs on FA metabolism have mostly focused on β-oxidation. This motivated us to investigate the elongation pathway of FAs after exposure to PFOSA, as part of a more global control of lipid synthesis and metabolism, and oxidative stress. Fisheries and aquaculture are economically important industries in several countries, and Atlantic salmon represent 90% of the production value in Norwegian aquaculture.24 Seafood has been found to highly influence human body burden of fluorinated compounds,25 underlining the importance of clarifying how environmentally ubiquitous PFCs influence both human and animal health and welfare. 1254

dx.doi.org/10.1021/tx300110u | Chem. Res. Toxicol. 2012, 25, 1253−1264

Chemical Research in Toxicology

Article

Table 1. Primer Pair Sequences, Accession Numbers, Amplicon Size, and Annealing Temperature Conditions for Genes of Interest Used for Real-Time PCR

a

Sequences are given in the 5′−3′order. A cell viability value of >90% was a criterion for further use of the cells. Cells were plated on 35 mm TPP Tissue Culture Plates (Techno Plastic Products AG, Switzerland) at a monolayer density of 2−2.4 × 106 cells in 3 mL of DMEM medium (without phenol red) containing 0.5% (v/v) FBS, 1% (v/v) L-glutamine, 15 mM HEPES, and 1% (v/v) antibiotic−antimycotic. The cells were cultured at 10 °C in a sterile incubator without O2/CO2 for 24 h prior to PFOSA exposure. 2.3. Exposure of Hepatocytes. After 24 h of preculture, the old medium was removed, and then 3 mL of fresh medium (without FBS) containing DMSO (solvent control) or 2, 20, and 50 μM PFOSA was added to the cells. The PFOSA was first dissolved in DMSO and then added to the medium. The final concentration of DMSO was 0.01% (v/v) in the exposure medium as well as in the solvent control. Because PFCs have been found to interact with serum albumin,27 no FBS was added to the exposure medium. In addition to the solvent control, an additional control with clean medium was applied as a treatment group. For the control, DMSO control, and low (2 μM), medium (20 μM), and high (50 μM) treatment groups, after 12 and 24 h, 1 mL of hepatocyte incubation media was collected in Eppendorf tubes and then snap frozen in liquid nitrogen. Samples (n = 6) for each treatment group were also harvested for total RNA isolation and lipid analysis. 2.4. Quantitative (Real-Time) PCR. Cells for total RNA isolation were harvested in Trizol according to the manufacturer's protocol (Invitrogen). The integrity of the RNA samples was verified by spectrophotometer analysis and formaldehyde agarose gel electrophoresis. Total cDNA for the gene expression analyses were generated from 1 μg of total RNA using a combination of poly-T and random hexamer primers from iScript cDNA Synthesis Kit as described by the manufacturer (Bio-Rad). High quality RNA with A260/A280 ratio above 1.9 and intact ribosomal 28S and 18S RNA bands was used for cDNA synthesis. Quantitative (real-time) PCR was used for evaluating gene expression profiles, and all genes were cloned and sequenced to confirm the gene product prior to using the primers. For each treatment, the expression of individual gene targets was analyzed using the Mx3000P Real-Time PCR System (Stratagene, La Jolla, CA, USA). Each 25-μL DNA amplification reaction contained 12.5 μL of iTAQSYBR Green Supermix with ROX (Bio-Rad), 1 μL of cDNA, and 200 nM of each forward and reverse primer (Table 1). The three-

2. MATERIALS AND METHODS 2.1. Chemicals and Reagents. The highly pure (>98%) linear perfluorooctane sulfonamide (PFOSA; CF3(CF2)7SO2NH2) isomer and isotopically labeled linear PFOSA-13C8 and linear PFOS-13C4 were purchased from Wellington Laboratories (Guelph, ON, Canada). iScriptTM cDNA Synthesis Kit and iTaqTM SYBR Green Supermix with ROX were supplied by BioRad Laboratories (Hercules, CA, USA). The original TA Cloning Kit PCR 2.1 vector, INVαF′ cells, TRIzol, and Dulbecco’s Modified Eagle's Medium (DMEM) with nonessential amino acid and without phenol red, fetal bovine serum (FBS), 0.4% trypan blue, and L-glutamine were purchased from GibcoInvitrogen Life Technologies (Carlsbad, CA, USA). Dimethyl sulfoxide (DMSO), penicillin−streptomycin−neomycin solution, collagenase (C0130-1G), bovine serum albumin (BSA), N-[2hydroxyethyl]piperazine-N′-[2-ethane sulfonic acid] (HEPES), ethylenediaminetetraacetic acid disodium salt dihydrate (EDTA), ethyleneglycol bis-(β-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA), and polyunsaturated fatty acids 1 and 2 (PUFA1 and PUFA2) were purchased from Sigma-Aldrich Chemie GmbH (Munich, Germany). Tricaine methane sulfonate (MS-222) was purchased from Norsk Medisinaldepot AS. GelRed Nucleic Acid Gel Stain was purchased from Biothium (Hayward, CA, USA). The ZR Plasmid MiniprepTMClassic was purchased from Zymo Research (Orange, CA, USA). 2.2. Collagenase Perfusion, Isolation, and Culture of Hepatocytes. Juvenile Atlantic salmon (Salmo salar: bodyweight of 120−150 g), were supplied by Marine Harvest AS, Norway. The experiment was performed in April/May 2010 on freshwater adapted individuals. The fish were kept at the animal holding facilities at Sealab, NTNU, where they were supplied with continuously running freshwater at a constant temperature of 10 °C. Prior to liver perfusion, all glassware and instruments were autoclaved, and solutions were filtration sterilized by using 0.22 μm Millipore filter (Millipore AS, Oslo, Norway). Hepatocytes were isolated by a two-step perfusion technique with modifications as previously described.26 The cell suspension was filtered through a 150 μM nylon monofilament filter and centrifuged at 70g for 5 min. Cells were washed three times with serum-containing medium and finally resuspended in complete medium. Following collagenase perfusion and isolation of hepatocytes, viability of cells was determined by the trypan blue exclusion method. 1255

dx.doi.org/10.1021/tx300110u | Chem. Res. Toxicol. 2012, 25, 1253−1264

Chemical Research in Toxicology

Article

Table 2. Hepatocellular Composition of Fatty Acids Classes (Expressed As the Percentage (%) of Total Fatty Acids) in Atlantic Salmon Hepatocytes Exposed for 12 and 24 h to Perfluorooctane Sulfonamide (PFOSA) at 2, 20, and 50 μMa

fatty acid myristic pentadecanoic palmitic palmitoleic heptadecanoic stearic oleic + elaidic linolelaidic γ-linolenic α-linolenic arachidic cis-11-eicosenoic cis-11,14eicosadienoic arachidonic cis-5,8,11,14,17-EPA heneicosanoic erucic cis-13,16docosadienoic lignoceric

12 h

24 h

[PFOSA, μM]

[PFOSA, μM]

control

2

20

50

control

2

20

50

14:0 15:0 16:0 16:1 17:0 18:0 18:1n9c+18:1n9t 18:2n6t 18:3n6 18:3n3 ALA 20:0 20:1n9 20:2

100 100 100 100 100 100 100 100 100 100 100 100 100

± ± ± ± ± ± ± ± ± ± ± ± ±

12 24 9 15 24 13 22 16 11 21 11 27 21

70 ± 12 75 ± 21 81 ± 13 76 ± 8 70 ± 14 51 ± 20* 93 ± 11 23 ± 7* 204 ± 55* 87 ± 23 12 ± 2* 96 ± 14 35 ± 17*

23 ± 4* 77 ± 12 39 ± 9* 42 ± 10* 115 ± 25 104 ± 21 52 ± 12* 13 ± 3* 251 ± 58* 80 ± 22 15 ± 3* 57 ± 14* 63 ± 22*

86 ± 17 99 ± 21 125 ± 21 109 ± 18 106 ± 23 113 ± 14 124 ± 22 100 ± 24 101 ± 13 99 ± 15 129 ± 31 130 ± 23 59 ± 12*

100 100 100 100 100 100 100 100 100 100 100 100 100

± ± ± ± ± ± ± ± ± ± ± ± ±

33 12 23 14 16 22 16 21 9 11 8 31 14

89 ± 11 80 ± 23 93 ± 16 99 ± 31 77 ± 11 73 ± 16 151 ± 32* 20 ± 4* 20 ± 3* 188 ± 35* 24 ± 6* 229 ± 63* 39 ± 9*

128 ± 28 200 ± 32* 49 ± 11* 171 ± 34* 108 ± 18 71 ± 11 155 ± 21* 5 ± 1* 246 ± 52* 109 ± 21 61 ± 10* 247 ± 62* 7 ± 2*

224 ± 51* 277 ± 59* 201 ± 62* 299 ± 45* 225 ± 37* 279 ± 66* 244 ± 57* 27 ± 9* 248 ± 54* 219 ± 47* 98 ± 21 258 ± 62* 16 ± 3*

20:4n6 ARA 20:5n3 EPA 21:0 22:1n9 22:2

100 100 100 100 100

± ± ± ± ±

10 6 12 25 21

31 ± 8* 126 ± 28 21 ± 4* 181 ± 31* 60 ± 17*

38 ± 9* 267 ± 14* 27 ± 4* 96 ± 16 64 ± 14*

57 ± 11* 185 ± 33* 44 ± 9* 97 ± 12 54 ± 11*

100 100 100 100 100

± ± ± ± ±

22 25 12 27 10

137 ± 33 294 ± 91* 125 ± 21 73 ± 14 128 ± 22

212 ± 55* 264 ± 61* 173 ± 32* 42 ± 10* 71 ± 13*

316 ± 78* 336 ± 98* 258 ± 72* 253 ± 75* 70 ± 13*

24:0

100 ± 8

26 ± 6*

59 ± 13*

120 ± 30

100 ± 25

14 ± 3*

35 ± 8*

187 ± 31*

a

Fatty acid levels were measured using gas chromatography−mass spectroscopy (GC-MS). Asterisks (*) denotes significant difference compared with control for individual time points. step real-time PCR program included an enzyme activation step at 95 °C (5 min) and 40 cycles of 95 °C (30 s), 60 °C (30 s), and 72 °C (30 s). Controls lacking a cDNA template were included to determine the specificity of target cDNA amplification. Cycle threshold (Ct) values obtained were converted into mRNA copy number using standard plots of Ct versus log copy number.28 The criterion for using the standard curve is based on equal amplification efficiency (usually >90%) with unknown samples, and this is checked prior to extrapolating unknown samples to the standard curve. The standard plots were generated for each target sequence using known amounts of linear plasmids containing the amplicon of interest. Data obtained from n = 6 samples were averaged and expressed as a percentage of control samples. 2.5. Glutathione S-Transferase (GST) Assay. GST activity in postmitochondrial supernatant (PMS) samples was determined in the present study by conjugation of reduced GSH to the aromatic substrate, 1-chloro-2,4-dinitrobenzene (CDNB), using the method of Habig et al.29 PMS was prepared from the cultured hepatocytes by centrifugation as previously described.30 During the procedure, the cells were constantly kept on ice to prevent protein degradation. The cells were centrifuged at 1000g for 3 min at 4 °C to sediment the cells without rupturing. The cells were immediately mixed with 200 μL of PMS buffer (pH 7.6) containing DTT (1 mM), NaH2PO4·H2O (0.1 M), KCl (0.15 M), EDTA, and glycerol (10 °C) and briefly homogenized by a Potter−Elvehjem type Teflon glass homogenizer (Glas-Col, Terre Haute, USA). The homogenate was centrifuged at 12000g for 20 min at 4 °C. The pellets were discarded, and PMS samples were collected and used for the analysis. A total of 30 μL PMS was applied to the wells of a transparent 96 well ELISA plate. Thirty microliters of PMS-buffer with DTT was used as control. Reduced GSH (1 mM) in CDNB buffer was prepared directly before use, and 200 μL of this solution was added to the PMS to start the reaction. The fluorescence was read at excitation and emission wavelengths of 535 and 590 nm for 20 min with the Synergi HT multidetection microplate reader. All the enzyme activities were assayed at room temperature (25 °C). The fluorescence data obtained within each exposure group was averaged and used to calculate the

enzyme activity by a preread standard curve. The results were normalized to the total protein content in each sample and expressed as nmol/min/mg protein. The total amount of PMS protein was determined by the method of Bradley (1976), using BSA standards. 2.6. FA Extraction and GC-MS Analysis. Lipids were extracted from the Atlantic salmon hepatocytes by homogenization in chloroform/methanol (2:1) solution, added with 0.01% of 2,6-ditert-butyl-4-methylphenol (BHT) as an antioxidant, according to the method of Folch et al.31 FA methyl esters (FAMEs) from total lipids were prepared by acid-catalyzed transmethylation for 1 h at 100 °C, using tricosanoic acid (23:0) as the internal standard. Methyl esters were extracted by c-hexane, then dried by centrivap, weighted, and suspended in c-hexane (1% v/v). The FAMEs analysis was performed using a Shimadzu GC-MS 2010 gas chromatograph−mass spectrometer and fitted with a fused silica capillary column (Supelco, Germany); helium was used as carrier gas. The injector, detector, and column temperatures were 250, 300, and 200 °C, respectively. The relative percentage of the area was obtained by using the following equation: Area % FAX = [AX/AR] × 100, where FAX = fatty acid to be quantified, AX = area of the methyl esters, X and AR = total area of the chromatogram. Peak areas lower than 0.1% of the total area was not considered. We identified FA methyl esters by comparing the retention time of the samples and standards. 2.7. PFOSA and PFOS Isolation from Samples. The incubation media were collected (1 mL) for each of the control, DMSO control, and low (2 μM), medium (20 μM) and high (50 μM) treatment group samples after 12 and 24 h, snap frozen, and shipped to Ottawa, Canada. All media samples were analyzed for PFOS and PFOSA in the Letcher/Organic Contaminants Research Laboratory (OCRL) at the National Wildlife Research Centre, Environnment Canada, Carleton University in Ottawa, Canada. As described in Chu and Letcher,32 PFOSA and PFOS were extracted directly from the isolated incubation media of the hepatocytes after the exposure period. Briefly, 100 μL of the media samples were diluted with 100 μL methanol solution of PFOSA-13C8 and PFOS-13C4 (100 ng/mL each) and an additional 100 μL of MeOH. The final solvent ratio in the HPLC vial was 2:1 1256

dx.doi.org/10.1021/tx300110u | Chem. Res. Toxicol. 2012, 25, 1253−1264

Chemical Research in Toxicology

Article

Figure 2. Modulation of fatty acid elongase (FAE: A), Δ5-desaturase (FAD5: B), and Δ6-desaturase (FAD6: C) mRNA in salmon hepatocytes exposed for 12 and 24 h to PFOSA at 2, 20, and 50 μM. mRNA (mRNA) levels were quantified using real-time PCR with gene specific primer pairs. The data are presented as mean values expressed as the percentage (%) of solvent control ± standard error of mean (SEM: n = 6). Asterisks denote significant difference compared to control, and different letters denote group means that are significantly different analyzed using ANOVA and the Holm−Sidak multiple comparison test (p < 0.05). cone gas flow rates were 600 and 100 L/h, respectively, while the desolvation and source temperatures were 350 and 120 °C, respectively. Argon was used as collision gas at 2.3 × 10−3 mbar pressure. Quantitative analysis was performed by the isotope dilution method based on the PFOSA-13C8 and PFOS-13C4 internal standards. A five point calibration curve was made spanning the range of anticipated analyte concentrations in the samples with fixed concentrations (33.3 ng/mL) of internal standards. A linear regression mode was selected on Masslynx V 4.0 instrument software for fitting of the calibration curve. Quantification was performed using an internal standard approach. Since an isotope dilution quantification approach was used, the concentrations were inherently recovery-corrected. 2.9. Quality Control and Data Analysis. The recovery efficiency of all the PFOSA and PFOS internal standards was generally greater than 77%. For every block of samples, a method blank of PFOSA and PFOS spiked distilled water was analyzed to assess the reproducibility of the method.32 For both PFOSA and PFOS, good reproducibility was obtained with a RSD of 8% (n = 10). 2.10. Statistical Analysis. The statistical analyses were performed with SigmaPlot, version 11.0. The analysis of variance (one-way

MeOH/aqueous phase, roughly the same concentration as the initial ratio of the HPLC mobile phase gradient. 2.8. Liquid Chromatography−Mass Spectrometry Analysis of PFOS and PFOSA. As described in Chu and Letcher,32 PFOS and PFOSA in the diluted incubation media samples were determined using a Waters Alliance 2695 HPLC system coupled to a Micromass Quattro Ultima triple quadrupole mass spectrometer equipped with electrospray ionization (ESI in the negative mode) interface was used for determination. The data was processed using Masslynx software (v 4.0). The analytes were separated chromatographically on an ACE 3 C18 column (50 × 2.1 mm i.d., 3 μm particle size). A sample volume of 10 μL was injected using an autoinjector. Separation was performed at 40 °C with a binary mixture of water (A) and methanol (B) as mobile phases both containing equal concentrations of 2 mM ammonium acetate. The elution gradient was as follows: initial mix of 40% mobile phase A and 60% mobile phase B, increasing to 100% mobile phase B within 10 min and held for 10 min, then decreasing to 60% mobile phase B within 1 min and held for 10 min. The MS/MS was operated in the multiple reaction monitoring (MRM) mode. The potential of the electrospray needle is held at 1 kV. The nebulizing and 1257

dx.doi.org/10.1021/tx300110u | Chem. Res. Toxicol. 2012, 25, 1253−1264

Chemical Research in Toxicology

Article

Figure 3. Modulation of peroxisome proliferator-activated receptors (PPARα, A; PPARβ, B; and PPARγ, C) and acyl-coenzyme A oxidase 1 (ACOX1: C) mRNA in salmon hepatocytes exposed for 12 and 24 h to PFOSA at 2, 20, and 50 μM. mRNA (mRNA) levels were quantified using real-time PCR with gene specific primer pairs. The data are presented as mean values expressed as the percentage (%) of solvent control ± standard error of mean (SEM: n = 6). Asterisks denote significant difference compared to the control, and different letters denote group means that are significantly different analyzed using ANOVA and Holm−Sidak multiple comparison test (p < 0.05). ANOVA) was used to measure statistical differences among the treatment groups. Significant differences between controls and different exposure groups were determined by the Holm−Sidak multiple comparison test after testing for normality and variance homogeneity. The level of statistical significance was set to p = 0.05.

3.2. Modulation of Hepatocellular FA Levels and Elongation. The hepatocellular FA profiles in control and PFOSA exposed cells are shown in Table 2. After 12 h, we observed significant increases in 18:3n-6 and 22:1n9 and decreases in 18:0, 18:2n6t, 20:0, 20:2, ARA, 21:0, 22:2, and 24:0 FAs in cells exposed to 2 μM. In the 20 μM PFOSA exposed cells, significant increases in 18:3n-6 and 20:5n-3 (EPA) and decreases in 14:0, 16:0, 16:1, 18:1n-9c+18:1n-9t, 18:2n6t, 20:0, 20:1n-9, 20:2, ARA, 21:0, 22:2, and 24:0 FAs were observed. Exposure of cells to 50 μM PFOSA for 12 h, produced significant increase in EPA, and decreases in 20:2, ARA, 21:0, and 22:2. The effects of PFOSA on lipids were more pronounced after 24 h of exposure, showing PFOSAdependent significant increases in most FA classes. In the 2 μM PFOSA exposed cells, we observed significant increases in 18:1n-9c+18:1 n-9t, ALA, and 20:1n-9, EPA and decreases in 18:2 n6t, 18:3n-6, 20:0, 20:2, and 24:0 FAs. For the 20 μM PFOSA exposed cells, significant increases in 15:0, 16:1, 18:1n9c+18:1n-9t, 18:3n-6, 20:1n-9, ARA, EPA, 21:0, decreases in 16:0, 18:2n6t, 20:0, 20:2, 22:1n-9, 22:2, 24:0 FAs were observed. Exposure of cells to 50 μM PFOSA produced

3. RESULTS 3.1. PFOSA and PFOS Analysis in Incubation Media. For the control and DMSO control, the concentrations of PFOSA and PFOS were nondetectable and