Nitro-Oxidative Stress and Mitochondrial Dysfunction in Human Cell Lines Exposed to the Environmental Contaminants PFOA and BPA

Background : Bisphenol A (BPA) and perfluorooctanoic acid (PFOA) are synthetic compounds widely utilized in industrial activities devoted to the production of daily life plastic, metal products, and packaging from which they are able to migrate to food and water. Due to their persistence in the environment, living organisms are chronically exposed to these pollutants. BPA and PFOA have adverse effects on tissues and organs. The aim of this study was to identify the molecular targets and biochemical mechanisms involved in their toxicity. Methods : HepG2 and HaCaT cells were treated with BPA or PFOA, and the trypan blue exclusion test and 3-(4,5-Dimethylthiazol-2-yl)- 2,5-Diphenyltetrazolium Bromide (MTT) assay were performed to define the conditions for subsequent investigations. We conducted quantitative PCR and western blot analysis to evaluate the expression of proteins involved in nitric oxide (NO) signaling. Cell-based assays were carried out to evaluate reactive oxygen species (ROS) production, nitrite/nitrate (NOx) accumulation, 3-nitrotyrosine (3-NT) formation, and mitochondrial membrane potential (MMP) determination in treated cells. Results : HepG2 and HaCaT cells incubated for 24 h with subtoxic concentrations of BPA or PFOA (50 and 10 µ M, respectively) exhibited altered mRNA and protein expression levels of NO synthase isoforms, manganese superoxide dismutase, and cytochrome c. Treatment with PFOA led to activation of inducible NO synthase (NOS), a marker of nitrosative stress, accompanied by the increased production of ROS, NOx, and 3-NT and alterations of the MMP compared to controls. Conclusions : The results of this study indicate the major involvement of the NO signaling axis in the persistent alteration of cell redox homeostasis and mitochondrial dysfunction induced by BPA and PFOA, highlighting the specific role of PFOA in NOS regulation and induction of nitro-oxidative stress.


Introduction
In healthy cells, the redox balance is maintained through a complex network of adaptive responses that include the continuous generation and elimination of reactive oxygen/nitrogen species (ROS/RNS). The fine-tune regulation of redox signaling events is activated by incoming insults (physical, chemical, or biological) to restore the physiological equilibrium [1][2][3][4][5].
Mitochondria are one of the main endogenous sources of ROS due to the formation of superoxide ion (O 2 −. ) at the level of the electron transport chain [6]. In addition, nitric oxide (NOꞏ) production is achieved through the specific activity of NO synthase (NOS) isozymes, i.e., neuronal NOS (nNOS), endothelial NOS (eNOS), and inducible NOS (iNOS) [7,8]. The nNOS and eNOS enzymes catalyze the production of NO needed for signaling [9,10]. By contrast, iNOS activation, characterized by a strong in-crease in NO concentration (µM), is associated with host defense mechanisms and immune responses [11], but is also responsible for deleterious effects such as the persistent inhibition of oxidative phosphorylation (OXPHOS), redox imbalance, and onset of inflammatory pathways [12][13][14][15].
The short-lived, highly reactive peroxynitrite (ONOO − ) is generated by the reaction of NO with O 2 − under conditions of increased bioavailability of these substrates. ONOO − is responsible for the direct or indirect oxidation of biological molecules [16] and for protein modifications occurring at the level of sensitive amino acids, such as the nitrosylation of cysteine and nitration of tyrosine residues [17,18].
A number of xenobiotics and related chemicals (e.g., atmospheric pollutants and industrial contaminants) are involved in the production of intracellular ROS/RNS, and the persistence of a given prooxidant stimulus may specifically affect redox homeostasis leading to chronic oxidative stress (OS) [4,15].
Diffuse OS is a common alteration in several pathological conditions including hypertension, diabetes, neurodegeneration and cancer [1,6,19,20]. Among the exogenous sources of nitro-OS, there has been growing interest in the chemical derivatives of synthetic compounds utilized in industrial manufacture. These chemicals, widely utilized as plasticizer and waterproof materials for packaging, utensils and pharmaceuticals, are now classified as emerging contaminants (ECs) as they are dispersed in the environment at alarming concentrations [21,22].
Bisphenol A (BPA) and perfluoro-octanoic acid (PFOA) are ideal representatives of this class of pollutants. Due to their intrinsic high stability, resistance to environmental conditions and long half-life, both have been widely used as components of utensils and personal care tools to confer durability, plasticity and impermeabilization [23][24][25].
The documented migration of BPA and PFOA from products to soil, water and food, and the consequent bioaccumulation of these ECs have established a condition of human chronic exposure to these pollutants through ingestion, inhalation and dermal absorption [26][27][28][29]. Great efforts are being made to clarify the adverse effects that both BPA and PFOA may have on living organisms, and to better understand the mechanisms underlying their toxic effects [27,28,[30][31][32][33][34].
Both BPA and PFOA are thought to target the mitochondria of exposed cells, thus inducing mitochondrial dysfunction and promoting an increase of both ROS and RNS with consequent onset of OS [39][40][41][42][43]. Glutamatergic neurons have increased expression of both iNOS and nNOS, resulting in the elevation of ROS and RNS and 3nitrotyrosine (3-NT) production induced by chronic exposure to BPA [44], whereas 24 h PFOA exposure results in increased ROS and antioxidant enzyme expression in the HepG2 human hepatoma cell line [45]. The proposed mechanisms of BPA and PFOA toxicities are linked to their structural and chemical properties, which favor cell permeability and point to the interactions with hormone receptors (i.e., estrogen receptor and peroxisome proliferator activated receptor alpha, respectively) in exposed cells [46][47][48]. Recent studies showed that PFOA was able to interact with superoxide dismutase (SOD) directly, inducing OS and apoptosis [49], and may lead to the production of proinflammatory cytokines, ROS/RNS increase and mitochondrial dysfunction in rat cell model [40].
In this work, the mechanisms of toxicity of the EDs BPA and PFOA were investigated, focusing on their ability to interfere with the NO signaling. To this end, two human cell models were chosen to assess the effects of BPA and PFOA treatments: the HepG2 hepatocellular cancerderived cell line, which is the gold standard model for xenobiotic metabolism and cytotoxicity studies; and the HaCaT immortalized keratinocyte cell line, which is a model system to study the skin as the outer barrier to environmental cytotoxic and genotoxic agents [50][51][52].
The expression of regulators involved in NO metabolism such as the NOS isoforms (eNOS in HepG2, nNOS in HaCaT and iNOS in both cell lines) was assessed at both the mRNA and protein levels, in association with evaluation of the expression of the mitochondrial proteins manganese SOD (MnSOD) and cytochrome c (cyt c). As a relevant parameter for the analyses of oxidative and nitrosative stress (NS), the level of ROS production in cells treated with BPA and PFOA was determined, as well as the accumulation of nitrites/nitrates (NOx), reflecting the increased production of NO compared to untreated cells. Furthermore, as a downstream effect of altered ROS/RNS homeostasis, we evaluated the level of 3-NT modification in proteins and alterations of the mitochondrial membrane potential (MMP).
Upon cell treatment with low µM concentrations of these substances (50 µM BPA and 10 µM PFOA) the physiologic cell redox homeostasis is affected, with mitochondrial alterations compatible with the onset of a persistent nitro-OS condition.
As members of a wide class of synthetic pollutants (ECs, EDs), both BPA and PFOA have analogues for which a comparable level of toxicity has been reported [27,28,38,39,49]. Further work is needed to clarify the effective EC pressure to which living organisms are exposed and related redox dysregulation is generated. We postulate that a comprehensive understanding of the specific contribution of abundant ECs (e.g., BPA or PFOA) to alterations in NO signaling and the promotion of OS and NS will provide a concrete basis by which to define the harmful effects of these pollutants. Of relevance, the major involvement of NO signaling in the toxicity exerted by the ECs BPA and PFOA and a novel remarkable effect of PFOA on the regulation of NOS isoforms were observed in this study. Within the limitation of the model system (cell lines) and application of selective incubation conditions, which do not represent the complexity of the whole organism, we found this cell line approach as an opportunity to evaluate the parameters related to altered ROS and RNS bioavailability.
The results of this study may provide insight into the effects derived from an impaired antioxidant cell response, which leads to an ROS/RNS increase, bioenergetic dysfunction and inflammation at the tissue level, a condition favoring the onset of pathological states [1,20,53].
Cells were incubated in the presence or absence of BPA and PFOA at different concentrations and times. When necessary, cells were harvested by trypsinization and centrifugation (1000 × g) and carefully suspended in working medium at a density of~2.8 × 10 4 cells/cm 2 (HepG2) and~1.2 × 10 4 cells/cm 2 (HaCaT). Cells were lysed with Cell Lysis Reagent (CelLytic TM M; Merck Life Science, Darmstadt, Germany) containing protease inhibitor cocktail (Roche, Basel, Switzerland). The BCA assay was used to determine the protein content.

Chemicals
DMEM and FBS were from Invitrogen Life Technologies (Gibco, Paisley, UK) and PAA Laboratories (Linz, Austria). BPA, perfluorooctanoic acid, JC-1, MTT, and trypan blue solution were from Merck Life Science. PFOA was dissolved in sterile dimethyl sulfoxide (DMSO; Merck Life Science) and further diluted in water; the administration of 10 µM PFOA led to residual DMSO <0.01%. BPA was dissolved in 10% ethanol; the administration of 50 µM BPA led to residual ethanol <0.01%. In the doseresponse experiments ( Fig. 1) vehicle (0.1% DMSO and 0.02% ethanol) was assayed corresponding to the highest volume of PFOA and BPA administered. Lipopolisaccaride (LPS) from Escherichia coli and interferon-γ (IFN-γ) were purchased from Merck Life Science. Nigericin and valinomycin utilized in the MMP determinations were from Abcam (Cambridge, UK).

Cell Viability Assays
The viability of HepG2 and HaCaT cells was assessed by the trypan blue exclusion test and MTT [3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazolium Bromide] reduction assay as previously described [54]. Briefly, cells seeded on 35 mm dish or 6-well plates were incubated for 24 h with BPA or PFOA (from 1 to 250 µM). After treatment, cells were detached with trypsin, pelleted at 800 × g, resuspended in phosphate-buffered saline (PBS) and mixed with equal volume of trypan blue. Live (bright clear) and dead (blue) cells were counted in the Thoma cell counting chamber. For the MTT assay, cells (2 × 10 5 /mL) were seeded in a 96-well plate in a final volume of 100 µL/well and in-cubated for 6 to 72 h at 37°C in the presence or absence of increasing BPA and PFOA concentrations (from 0.1 µM to 1 mM). Then 10 µL MTT solution (5 mg/mL) was added to each well, followed by a 4 h incubation at 37°C. To dissolve the dark-colored formazan crystals produced by reduction of the MTT tetrazolium salt, cells were incubated for 30 min at 37°C in the dark with 100 µL DMSO. The optical density of reduced MTT was measured at 570 nm with a reference wavelength at 690 nm using the Appliskan Microplate Reader (Thermo Fisher Scientific, Waltham, MA, USA).

Quantitative PCR
Quantitative PCR (qPCR) was carried out in HepG2 and HaCaT cells. Briefly, cells were incubated for 24 h with BPA (50 µM) or PFOA (10 µM). After incubation, cells were harvested by trypsinization and centrifugation (1000 × g for 5 min at 20°C), washed twice with Hank's buffer and counted. Cells (1 × 10 6 ) were lysed in 350 µL RA1 Buffer containing 10 mM DTT, and RNA was isolated and purified using the Nucleo Spin® RNA Kit (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions. Then, 1 µg total RNA was used for reverse transcription reaction with the RT2 First Strand Kit (Qiagen, Venlo, Netherlands) and qPCR was performed using primers designed by Bio-Rad Laboratories (Hercules, CA, USA) (Software Beacon Designer) and purchased by PrimmBiotech, Cambridge, USA. SYBR Green qPCR (Brilliant_SYBR_Green QPCR Master Mix) was performed using the Stratagene Mx3005p System (both from Agilent Technologies, Santa Clara, CA, USA). The cDNAs were amplified using 45 cycles consisting of a denaturation step (95°C for 5 min) and amplification step (95°C for 10 s, 55°C for 30 s). Melting curve analysis was performed at the end of every run to ensure the presence of a single amplified product for each reaction. The β-actin gene (PrimmBiotech) was used for normalization. The primers used were as follows. nNOS (NOS1) forward:

Western Blot Analysis
HepG2 and HaCaT cells (3 × 10 6 cells) were incubated for 24 h with BPA (50 µM), PFOA (10 µM) or untreated, followed by lysis with CellLytic TM M reagent in the presence of protease inhibitors (both from Merck Life Science). Cells were also incubated for 24 h with 1 mg/mL LPS and 10 ng/mL IFN-γ (both from Merck Life Science) as a positive control of iNOS induction. Protein samples (50 µg) were separated on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis gels and electrotransferred to polyvinylidene difluoride membranes. After blocking for 1.5 h in 4% bovine serum albumin, membranes were incubated overnight at 4°C with the following primary antibodies: polyclonal anti-iNOS antibody (Boster Biological Technology, Pleasanton, CA, USA), polyclonal anti-eNOS and anti-nNOS antibodies (both from Santa Cruz Biotechnology, Dallas, TX, USA), polyclonal anti-MnSOD antibody (Bio-Rad), monoclonal anti-cyt c antibody (Santa Cruz Biotechnology) and monoclonal anti β-actin antibody (Invitrogen, Waltham, MA, USA). All primary antibodies were diluted 1:1000 in blocking buffer, except anti βactin antibody that was diluted 1:5000. After washing with PBS/Tween, the membranes were incubated for 1 h at room temperature with horseradish peroxidase-conjugated secondary antibodies (1:5000; Merck Life Science). Proteins were detected with ChemiDocTM MP Image Analysis Software (Bio-Rad) after the addition of Enhanced Chemiluminescence Western Blotting Substrate (Bio-Rad). When necessary, membranes were stripped (in glycine/SDS/Tween 20, pH 2.2) and reprobed with anti-eNOS/nNOS antibodies. Densitometric analysis was carried out with Image Lab 6.0.1 software (Bio-Rad, CA, USA).

Reactive Oxygen Species Quantification
Reactive oxygen species (ROS) generation was assessed in living cells using two fluorescent probes with a slight difference in ROS specificity for independent evaluations: red-fluorescent MAK145 (Merck Life Science) and 2 ′ ,7 ′ -dichlorodihydrofluorescein diacetate (DCFDA) (Cayman Chemical Company, Ann Arbor, MI, USA). Before the assay, HepG2 and HaCaT (~1 × 10 6 cells/mL) were incubated for 24 h in a 24-well (black) plate in the presence of BPA 50 µM and PFOA 10 µM or untreated; cells were also assayed in the presence of H 2 O 2 (0.1 mM) as a positive control. MAK145 was added 1 h before the termination of treatments. After 1 h of incubation, according to the manufacturer's instructions, the fluorescence intensity was measured at λ ex = 520 and λ em = 605 nm. DCFDA (10 µM) was incubated for 30 min at 37°C after treatment, and the fluorescent signal was measured at λ ex = 485 and λ em = 535 nm. Fluorescence was measured with the VICTOR TM Multilabel Counter Plate Reader (Perkin Elmer, Waltham, MA, USA).

NOx Determination
The total NOx was evaluated as a quantitative measure of NO production. The accumulation of NOx was assessed in the culture medium of HepG2 and HaCaT cells (~2.5 × 10 5 cells/mL), which were grown overnight in DMEM 1 g/L glucose (without FBS and phenol red), upon 24 h exposure to BPA 50 µM and PFOA 10 µM. After incubation, the cell supernatants were centrifuged at 1000 × g for 10 min at 4°C, and the NOx content was determined fluorometrically using the fluorescent probe 2, 3-diaminonaphthalene (DAN) (Nitrate/Nitrite Fluorometric Assay Kit; Cayman Chemical Company). The fluorescent intensity, which is proportional to total NO production, was measured with the Fluorescence Plate Reader VICTOR Multilabel Counter (Perkin Elmer, Waltham, USA) using an excitation wavelength of 375 nm and emission wavelength of 420 nm.

Detection of 3-NT Protein Modifications
The evaluation of 3-NT modified proteins was used as marker of peroxynitrite-mediated NS in HepG2 and Ha-CaT cells treated with BPA 50 µM and PFOA 10 µM or untreated. After treatments, HepG2 cells were trypsinized, centrifuged at 500 × g for 10 min and washed twice with PBS. The cell pellet (1 × 10 6 cells) was resuspended with extraction buffer and incubated on ice for 20 min. After centrifugation at 12000 × g 4°C for 20 min, the 3-NT levels were assessed colorimetrically using the competitive Nitrotyrosine ELISA Kit (Abcam) and normalized to the total protein content determined by the BCA assay. The absorbance was measured at 450 nm using the Appliskan Microplate Reader (Thermo Fisher Scientific).

Statistical Analyses
Data are reported as the mean ± standard deviation of at least three independent experiments and significance (p values) were determined using the unpaired Student's t-test. p ≤ 0.05 was considered statistically significant.

Cytotoxicity of BPA and PFOA in HepG2 and HaCaT cells
The cytotoxicity of BPA and PFOA in HepG2 and HaCaT cells was determined by the trypan blue exclusion and MTT assays (Fig. 1). Dose-response analyses were performed by incubating both cell lines with increasing amounts of BPA or PFOA. As shown in Fig. 1A-C, 24 h exposure to BPA induced similar effects on HaCaT and HepG2 cells, namely, a decrease in cell viability of about 20% in a concentration range of 50-100 µM, followed by an additional decrease in viability of about 50% at a BPA concentration of 250 µM. After further increasing the BPA concentration, higher toxicity was observed, leading to residual viability of~20% in both HaCaT and HepG2 cell lines. Dose-response analyses after the administration of PFOA revealed a 20% decrease in the viability of both cell lines at 10 µM, followed by an additional decrease at 50 µM PFOA (~70% overall viability) (Fig. 1B,D). In a concentration range of 100-250 µM PFOA, the overall decreasing trend observed was more pronounced for HaCaT than HepG2 cells. To better analyze the sensitivity of HaCaT and HepG2 cells to BPA/PFOA and to identify the condition of sublethal toxicity suitable for further investigations, the MTT assay was conducted for the time-dependent evaluation of the effects exerted by the ECs (Fig. 1G,H). At a concentration of 50 µM for BPA or 10 µM for PFOA, the overall cell metabolic activity was within a 20% value, and a decrease in cell viability was observed as a function of time. Extended exposure to BPA of both cell lines (Fig. 1E) resulted in a further~10% decrease of metabolically active cells, reaching 70% after 72 h. A similar trend was observed after extending the PFOA incubation time, and slightly higher sensitivity was observed in HaCaT cells compared to HepG2 cells, with the former displaying a marked decrease in cell viability (30% viable cells) compared to the latter (60% viable cells) after 72 h of PFOA treatment (Fig. 1F).
The results of the viability assays allowed us to identify a suitable concentration and incubation time for further experiments: 50 µM for BPA or 10 µM PFOA, for 24 h. These conditions were indeed considered subtoxic, accounting for the limited metabolic alteration observed (within 20%).

NOS Isoform Expression in HepG2 and HaCaT Cells Exposed to BPA and to PFOA
The mRNA and protein levels of the different isoforms of NOS (iNOS and tissue-specific NOS, eNOS produced in hepatocytes [HepG2], and nNOS expressed in keratinocytes [HaCaT]) were determined after incubation with BPA or PFOA for 24 h. Compared to controls, BPA induced a two-fold increase in both iNOS (p = 0.015 vs ctr) and eNOS (p = 0.023 vs ctr) mRNA in HepG2 cells; the effect of PFOA on iNOS activation was higher (four-fold; p = 6 × 10 −4 vs ctr), but differently from BPA, PFOA weakly decreased eNOS mRNA (less than 10%) (p = 0.0022 vs ctr) (Fig. 2A). In parallel, in HaCaT cells, BPA treatment resulted in increased mRNA expression of both NOS isoforms (Fig. 2B), but the effect was limited to a 1.5-fold increase (p = iNOS 0.054; p = nNOS 0.016 in HaCaT cells) compared to HepG2 cells. Similar to that observed in HepG2 cells, PFOA induced a significant decrease in nNOS mRNA in keratinocytes (~25%; p = 0.047 vs ctr), whereas iNOS was significantly upregulated (~1.4-fold; p = 0.017 vs ctr). The protein expression of the two NOS isoforms in HepG2 cells was consistent with the changes in mRNA expression, with a maximum increase of 50% upon PFOA treatment (p iNOS BPA = 0.026 vs ctr; p eNOS BPA = 0.013 BPA vs ctr; p iNOS PFOA = 6.1 ± 10 −5 vs ctr); p eNOS = 0.013 BPA vs ctr) compared to controls (Fig. 2C). A similar increase with respect to iNOS expression was observed after 24 h cell treatment with a mixture of IFN-γ and LPS (p = 0.0067 vs ctr), a positive control of iNOS induction (Fig. 2C). NOS protein detected by Western blotting in Ha-CaT cells showed iNOS increase above 1.5-fold (p = 0.017 vs ctr) upon PFOA treatment, whereas the effect of BPA on iNOS induction was only very small although significant (~10%; p = 0.026 vs ctr) (Fig. 2E). nNOS production was found to be comparable with the control level (Fig. 2E) after the various treatments, except for a narrow but significant increase (20%; p = 0.014 vs ctr) induced by BPA (Fig. 2E). The effect of PFOA on NOS protein expression was reproducible in the cell line assayed, with iNOS induction as a specific target.

MnSOD and cyt c Expression in HepG2 and HaCaT Cells Exposed to BPA and to PFOA
The mRNA expression levels of the mitochondrial proteins MnSOD and cyt c were detected following treatment of HepG2 and HaCaT cells with BPA or PFOA under the conditions previously described. The level of Mn-SOD mRNA was not significantly altered in HepG2 after 24 h of BPA exposure, and was slightly induced by PFOA (1.25-fold; p = 0.014 vs ctr) (Fig. 3A), although at the protein level, the significant upregulation of MnSOD was induced by both EDs (40% BPA increase; p BPA = 0.02 vs ctr; 60% PFOA increase; p PFOA = 1.7 × 10 −5 vs ctr), as shown in Fig. 3B. In HaCaT cells, a~two-fold increment in Mn-SOD expression was induced by PFOA at mRNA level (p = 0.018 vs ctr) (Fig. 3C). Furthermore, the protein expression detected by anti-MnSOD antibody was increased by both BPA (40% increment; p = 0.019 vs ctr) and PFOA (34% in-crement; p = 2.1 × 10 −6 vs ctr) (Fig. 3D), similar to what was observed in HepG2 cells under comparable incubation conditions.
The analysis of cyt c expression showed mRNA induction specifically driven by PFOA treatment of both cell lines, with a stronger effect observed in HepG2 cells (above two-fold, p = 0.0036; Fig. 4A) compared to Ha-  CaT (~1.8-fold, p = 0.045; Fig. 4C). BPA treatment led to a small but significant decrease in cyt c mRNA (35% de-crease; p = 0.036), with the effect limited to HaCaT cells, as in HepG2 cells, there was no significant change in cyt c mRNA expression after BPA treatment compared to the control (Fig. 4A). As shown in Fig. 4B, cyt c protein expression detected by Western blot analysis in HepG2 cells was~1.4-fold higher in response to both BPA (p = 0.0087) and PFOA (p = 0.01) treatment, whereas the response of HaCaT cells regarding cyt c protein expression (Fig. 4D) was reproduced only in the case of PFOA treatment (~1.5fold increase; p = 0.0013) but not by BPA. In keratinocytes, indeed, BPA led to no changes in cyt c protein expression under the condition assayed.
Overall, in HaCaT cells, 24 h treatment with BPA showed a lower sensitivity in terms of cyt c and MnSOD expression with respect to HepG2 and compared with the effects exerted by PFOA.

ROS, NOx, and 3-NT Induction by BPA and PFOA
The production of ROS, the accumulation of NOx and the amount of NT modifications in proteins (3-NT) were measured to assess the OS and NS levels in HepG2 and HaCaT cells after 24 h treatment with BPA or PFOA. ROS determination was carried out taking advantage of two different ROS-targeting probes, for independent evaluations: DCFDA (Fig. 5A), mainly targeting hydrogen peroxide (H 2 O 2 ) but with broad sensitivity towards other ROS and RNS, such as peroxynitrite; and the MAK-145 redfluorescent probe (Fig. 5B) with selectivity for superoxide and hydroxyl radicals. Under the conditions assayed (24 h incubation with 50 µM BPA or 10 µM PFOA), ROS levels were specifically increased to~50% with respect to control values as revealed by the DCFDA assay (Fig. 5A). Specif-ically, BPA induced a~40% ROS increase in HaCaT (p = 0.006 vs ctr) but not HepG2 cells, whereas PFOA treatment induced a similar increase in both HepG2 (45% increase; p = 0.0117 vs ctr) and HaCaT (50% increase; p = 0.045 vs ctr) cells. IFN-γ + LPS and H 2 O 2 were used as positive controls. A much lower ROS level was detected by the MAK-145 probe, with significance in the case of BPA treatment of HepG2 cells (20%; p = 0.043 vs ctr) but not in HaCaT cells (p = 0.55 vs ctr). A small but significant ROS induction was also observed upon PFOA treatment of HepG2 (27%; p = 0.025) and HaCaT (32%; p = 0.041) cells compared to controls (Fig. 5B). The levels of NO end products, NOx, were found to be increased by treatment with both BPA and PFOA in the two cell lines tested (Fig. 5C). The effect was more evident in PFOA-treated cells, resulting in the increased accumulation of NOx by about two-fold and 1.5-fold for HepG2 (p = 0.0012) and HaCaT (p = 0.05) cells, respectively. A smaller but clearly significant induction of NOx was also achieved by BPA, by about 1.5-fold in both cell lines (p HepG2 = 0.009 and p HaCaT = 0.045 vs controls). It is worth noting that the result of increased NOx accumulation was in good agreement with the observation of increased NOS expression (especially iNOS) observed upon ED treatments.
As shown in Fig. 5D, the amount of NT modification in proteins was increased of about 25% in HepG2 cells following treatment of both BPA (p = 0.029 vs ctr) and PFOA (p = 0.018 vs ctr). Similar changes were seen in HaCaT cells exposed to PFOA (p = 0.021 vs ctr), whereas only a small irrelevant increase was observed for BPA (lower than 20%) under the assayed conditions. It is worth mentioning that a 24 h incubation of IFN-γ/LPS was able to specifically raise NO concentration in cells by iNOS activation, leading to the highest NT level, whereas the iNOS inhibitor 1400W showed the lowest level of NT, comparable in average to the control.

MMP of BPA-and PFOA-Treated HepG2 and HaCaT Cells
The overall mitochondrial functional state was investigated by measuring the MMP of BPA-or PFOA-treated cells, by evaluating the mitochondrial import of the fluorescent probe JC-1 and the formation of the red J-aggregates. As shown in Fig. 6A, after 24 h incubation with BPA, the MMP was significantly lowered in HepG2 (~20% de-crease; p = 0.0115) and HaCaT (~40% decrease; p = 0.023), whereas no significant difference was induced by PFOA. Surprisingly, the response of the MMP to the two-step addition of the ionophores nigericin and valinomycin, indicated a condition of hyperpolarization determined by the incubation of cells with PFOA. The hyperpolarization was significant in HepG2 (~30% decrease; p = 0.032) with respect to the polarization level of the untreated cells, whereas a small non-significant depolarization was observed for BPA treatment (Fig. 6B). The MMP alterations described, specifically induced by BPA or PFOA, were observed in the two cell lines assayed.

Discussion
Environmental diffusion of the ECs BPA and PFOA, compounds widely utilized in industrial manufacturing processes [21,22], has resulted in alarming concentrations detected in water and soil as well as bioaccumulation, as demonstrated by the detection of both BPA and PFOA in human body fluids with documented mother to fetus transfer [56,57]. The adverse effects exerted by chronic exposure of BPA and PFOA have been described over the last several years [27,32,35]. Of relevance, the ED action shown for both compounds [33,35,36], and related alterations of the physiological hormonal signals are strictly connected to the perturbation of cellular pathways related to redox homeostasis. The involvement of BPA in OS induction has been previously suggested based on its interaction with estrogen receptors, although the mechanisms of this process need to be elucidated [27,31,41,44,58].
This paper reports a detailed characterization of the biochemical parameters related to OS and NS induced by BPA, showing a correlation between alterations of ROS and RNS signaling and establishment of a cell redox imbalance with consequences at the mitochondrial level. Furthermore, the novelty of the specific involvement of PFOA in the perturbation of the NO signaling axis is proposed here, with activation of iNOS and production of peroxinitrite, suggesting a specific effect exerted by PFOA toward NS induction, under the condition assayed.
The limitations of the study with respect to the biomedical relevance of the findings must be taken into account, with particular reference to the cell models and the concentration of ECs assayed. As transformed/immortalized cells, the HepG2 and HaCaT cell lines utilized in this study may display altered metabolism and viability respect to normal cells. However, it must be considered that the cell lines utilized represent a model system for toxicological studies and that the experimental condition utilized (low glucose adaptation, FBS and phenol red starvation) were adopted to minimize eventual deviations in the results. The screening of dose-response and time dependency of the ECs carried out by the trypan blue exclusion test and MTT assay (Fig. 1) was conducted to establish a suitable concentration and time window for further investigations regarding OS and NS in cell models of interest.
At the concentrations of 50 µM BPA or 10 µM PFOA, only minor effects were observed on cell viability (≤20%; Fig. 1A-F), whereas the same doses led to specific induction of mRNA encoding for the iNOS enzyme, with stronger effects exerted by PFOA (four-fold iNOS induction in HepG2 cells) with respect to BPA (two-fold iNOS induction in HepG2 cells). BPA was able to induce a significant effect in terms of a "physiological" NOS (eNOS/nNOS) modulation, whereas PFOA failed to evoke this specific cell response ( Fig. 2A,B). These considerations can be extended to the protein level (Fig. 2C-F), as the increase in iNOS and e/nNOS showed a trend consistent with the mRNA data, although a smaller amplitude of the effect (1.6-fold average increase) was observed, and with a relevant exception given by the absence of iNOS induction in HaCaT cells by BPA. These observations strongly suggest that both BPA and PFOA have the ability to specifically affect the NO signaling axis, although through different mechanisms.
The increased expression of the "physiological" tissue-specific NOS isoform (eNOS in HepG2 and nNOS in HaCaT) induced by BPA may indicate the activation of adaptation pathways that possibly stimulated a specific cell stress response counteracting the detrimental activity of iNOS to limit the increase of highly reactive molecules to an acceptable range in terms of OS/NS. Under the condition assayed (50 µM, 24 h), in agreement with a previous observation, BPA incubation was associated with increases in MnSOD protein expression (Fig. 3). A parallel induction of cyt c expression was observed in HepG2 BPA-treated cells (Fig. 4), whereas HaCaT was insensitive to BPA with respect to cyt c stimulation. The variation in ROS level observed was generally small in amplitude (Fig. 5A,B), in accordance with an increase in MnSOD activity. In the case of BPA treatment, ROS were almost comparable with untreated cells, except for a 1.5-fold increase in HaCaT detected by DCFDA, which may possibly be associated with less evident MnSOD activation in that cell line (Fig. 3D). The increase in NO end products induced by BPA in both cell lines was in agreement with activation of the NO signaling axis (Fig. 5C). These results were in line with an increase in 3-NT protein modifications observed in BPAtreated HepG2 cells, although the results were not significant in HaCaT cells (Fig. 5D). In keratinocytes, the analysis of data arising from BPA treatments would seem indicative of their lower sensitization to this toxic compound; however, these results are possibly related to a cell response compatible with a different time window with respect to the condition assayed, as suggested by the increase in ROS production observed after 1-4 h treatment with BPA (not shown).
At the mitochondrial level, BPA administration led to a decrease in MMP in both cell lines (Fig. 6A). In HaCaT cells, this result was associated with the low involvement of cyt c with respect to BPA-induced modulatory events (Fig. 4C,D), revealing a depressed mitochondrial function. In view of the endocrine disruptive activity of BPA, under the conditions herein explored, we clearly showed that alteration of the NO axis [43,44] and dysregulation of the mitochondrial parameters [41], which have been previously proposed, are indeed strictly correlated and may be associated with the induction of nitro-OS [31,33].
The results observed upon PFOA treatment (10 µM, 24 h) showed the activation of iNOS at both the mRNA (four-fold increase in HepG2 cells) and protein (60% increase) levels. Differently from BPA, this effect was not associated to a parallel increase of the physiological NOS isoform, but iNOS induction was accompanied by a small but significant decrease in eNOS and nNOS mRNA in HepG2 and HaCaT cells, respectively ( Fig. 2A,B). The observation of the specific effect of PFOA on NO signaling is novel, and is related to the documented inverse "crosstalk" occurring among the NOS isoforms under specific cell conditions approaching the onset of NS [15]. The increase of iNOS expression induced by PFOA led the inducible isoform to prevail over eNOS, suggesting the onset of a more severe level of oxidative and nitrosative damage induced by PFOA with respect to BPA, especially in hepatocytes, where the effects of long-lasting NO increase derived by iNOS activity might trigger the initiation of inflammation. The modulatory trend among NOS isoforms observed at the mRNA level and confirmed at the protein expression level was associated with the increment in OS and NS markers, such as ROS production, NOx accumulation, and 3-NT protein modification (Fig. 5A-D). The results discussed above with respect to HepG2 cells exposed to PFOA were only partially reproduced in HaCaT cells, where the same treatment was able to only induce a more confined but still significant iNOS activation (Fig. 2B) confirmed by NOx accumulation, which was not associated with a substantial increase in 3-NT-specific damage. In both cell lines, PFOA treatment was able to induce a small but significant increase in MnSOD expression under the conditions assayed (Fig. 3), a result that, as mentioned above, may possibly explain the general finding of a limited increase in ROS levels compared to controls (Fig. 5A,B). However, the evidence of a consistent increment in DCFDA-targeted ROS detected in HepG2 cells might be due to the sensitivity of DCFDA for RNS and associated with an NS-specific induction driven by PFOA in HepG2 cells. Focusing on mitochondrial functional parameters, PFOA was able to induce a significant increase in mitochondrial potential, as indicated by the increased ∆ψ component of the MMP observed in the presence of ionophores (Fig. 6B). It is worth noting that the observed mitochondrial hyperpolarization specifically induced by PFOA was associated with increased cyt c expression (Fig. 4A,B) suggesting a degree of inhibition at the OXPHOS level. The results presented point to an association between alteration of mitochondrial function and OS, and highlight the role of the NS component generated through activation of the NO-axis to the overall degree of cell dysfunction induced by both EDs, particularly PFOA.

Conclusions
BPA and PFOA are both able to influence and alter the NO signaling pathway and promote the alteration of mitochondrial function. In hepatocytes, PFOA determines a proinflammatory condition, characterized by the high level of iNOS expression, increased ROS, significant alteration of proteins at the nitro-tyrosine level, and mitochondrial hyperpolarization. In the same cells, BPA was active in deregulating physiologic eNOS/iNOS crosstalk at the expression level, leading to elevation of NT modification with consequences at the mitochondrial level, as suggested by the decreased MMP. The keratinocyte response to insult arising from contaminant administration in the case of PFOA resulted in lower involvement of NO signaling compared to HepG2, but still with indications of an altered MMP and MnSOD activation at the mitochondrial level. The observed (mild) induction of iNOS mRNA driven by BPA was not accompanied by increased iNOS protein in keratinocytes, with the overall effect of a minor involvement of the NO axis under the tested conditions. However, BPA was able to induce persistent mitochondrial depolarization in keratinocytes, in accordance with the endocrinedisruption physiological alterations induced by this compound.
The experimental work herein presented shows that BPA and PFOA may target different components of the NO signaling pathway, with downstream effects mostly related to alteration of cell redox homeostasis. Further investigations are needed to fully understand the bioenergetic alterations and cell response linked to the nitro-OS induced by these pollutants. The findings reported will help clarify the mechanisms of action of BPA and PFOA, which will aid in the management of pathological relapses linked to chronic exposure to these environmental contaminants.

Author Contributions
MCM and MA designed the research study. MX, BS, GA and FD performed the research. MCM and BS analyzed the data. MA, MCM and BB wrote the manuscript. EP and LS provided advice on data finalization and supervised manuscript preparation. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

Ethics Approval and Consent to Participate
Not applicable.