Alterations of gene expression indicating effects on

0 downloads 0 Views 3MB Size Report
Nov 2, 2016 - a large amount of nutrients and organic matter (Degobbis et al.,. 2000 .... areas, A and B. Each transect contained 4 sampling stations located .... Cells were seeded on 24-well Falcon Primaria™ culture plates ...... cresa.it/_annuario/2016/Annuario2016.pdf. ... Richmond river estuary northern NSW, Australia.
Marine Environmental Research 123 (2017) 25e37

Contents lists available at ScienceDirect

Marine Environmental Research journal homepage: www.elsevier.com/locate/marenvrev

Alterations of gene expression indicating effects on estrogen signaling and lipid homeostasis in seabream hepatocytes exposed to extracts of seawater sampled from a coastal area of the central Adriatic Sea (Italy) Paolo Cocci a, Martina Capriotti a, Gilberto Mosconi a, Alessandra Campanelli b, Emanuela Frapiccini b, Mauro Marini b, Giovanni Caprioli c, Gianni Sagratini c, Graziano Aretusi d, e, Francesco Alessandro Palermo a, * a

School of Biosciences and Veterinary Medicine, University of Camerino, Via Gentile III Da Varano, I-62032 Camerino MC, Italy National Research Council, Institute of Marine Science CNR-ISMAR, L.go Fiera della Pesca, 2, 60125 Ancona, Italy c School of Pharmacy, University of Camerino, Via Sant'Agostino 1, I-62032 Camerino MC, Italy d Controllo Statistico, Pescara, Italy1 e Marine Protected Area Torre del Cerrano, 64025 Pineto, TE, Italy b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 21 June 2016 Received in revised form 5 October 2016 Accepted 1 November 2016 Available online 2 November 2016

Recent evidences suggest that the toxicological effects of endocrine disrupting chemicals (EDCs) involve multiple nuclear receptor-mediated pathways, including estrogen receptor (ER) and peroxisome proliferator-activated receptor (PPAR) signaling systems. Thus, our objective in this study was to detect the summated endocrine effects of EDCs with metabolic activity in coastal waters of the central Adriatic Sea by means of a toxicogenomic approach using seabream hepatocytes. Gene expression patterns were also correlated with seawater levels of polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs). We found that seawater extracts taken at certain areas induced gene expression profiles of ERa/vitellogenin, PPARa/Stearoyl-CoA desaturase 1A, cytochrome P4501A (CYP1A) and metallothionein. These increased levels of biomarkers responses correlated with spatial distribution of PAHs/ PCBs concentrations observed by chemical analysis in the different study areas. Collectively, our data give a snapshot of the presence of complex EDC mixtures that are able to perturb metabolic signaling in coastal marine waters. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Endocrine disruption Oestrogen mimics Gene transcription Coastal waters Adriatic Sea Sparus aurata

1. Introduction The Mediterranean Sea basin is subjected to a strong anthropic impact and is characterized by one of the most highest impacted coastal marine environment (Coll et al., 2010; Naser, 2013). As a distinct sub-region within the Mediterranean Sea, the North Adriatic's basin, a highly productive area of the Mediterranean (Campanelli et al., 2011; Grilli et al., 2013; Zavatarelli et al., 1998), is clearly more influenced by river floods (Cozzi and Giani, 2011; Djakovac et al., 2012; Marini et al., 2002) that affect both the circulation through buoyancy input and the ecosystem by introducing a large amount of nutrients and organic matter (Degobbis et al.,

* Corresponding author. E-mail address: [email protected] (F.A. Palermo). 1 http://www.controllostatistico.com http://dx.doi.org/10.1016/j.marenvres.2016.11.001 0141-1136/© 2016 Elsevier Ltd. All rights reserved.

2000; Marini et al., 2008, 2015). The Po River provides the major buoyancy flux with an annual mean freshwater discharge rate of 1500 m3 s1 (Cozzi and Giani, 2011; Raicich, 1996). The riverine waters discharging into the Northern Adriatic form a buoyant layer that typically flows southward along the Italian coasts and is constrained close to the coast over the continental shelf, more than 50 m deep (Poulain, 2001; Poulain and Cushman-Roisin, 2001). The northern and middle sub-basins of the Adriatic Sea have proven to be particularly sensitive to the accumulation of pollutants because of the considerable presence of rivers combined with limited recirculation and shallow depth. This is particularly evident when considering the western coast of the Adriatic Sea that is contaminated by sewage, industrial and agricultural outfall showing high levels of heavy metals, polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs) pollution (Annibaldi et al., 2015; Korlevic et al., 2015; Pavoni et al., 2003). To date, it has been amply demonstrated that most of these pollutants can induce

26

P. Cocci et al. / Marine Environmental Research 123 (2017) 25e37

toxicological effects to wildlife working as endocrine disrupting chemicals (EDCs) (Barber et al., 2000; Isidori et al., 2010; Tanabe, 2002). It is likely that dioxin-like PCB metabolites elicit estrogenic activity by stimulating estrogen receptor (ER) signaling in fish primary hepatocytes (Mortensen and Arukwe, 2008). Similarly, exposures to PAHs have been implicated in affecting reproductive parameters in juvenile fish (Spromberg and Meador, 2005). Several studies have reported the presence of EDCs in seawater, sediments and suspended solids along the Mediterranean sea area (GomezGutierrez et al., 2007; Micheletti et al., 2007; Pinto et al., 2005; Pojana et al., 2004). In addition, EDCs have been detected in a variety of molluscs, crustaceans and fishes from the Adriatic Sea (Ferrara et al., 2001, 2005). However, despite the many EDCs with the potential to contaminate the marine environment, biomonitoring data on the levels of these hazardous substances into the Adriatic Sea area are limited. Available evidences from aquatic environment studies suggest that the toxicological effects of EDCs involve multiple nuclear receptor-mediated pathways, including ER and peroxisome proliferator-activated receptor (PPAR) signaling systems (Bemanian et al., 2004; Fang et al., 2012; Ishibashi et al., 2008; Moody et al., 2002; Pavlikova et al., 2010). The detection of the estradiol-17b (E2)-inducible protein vitellogenin (VTG) in different species of teleost fish has been successfully utilized for detecting exposure to xenoestrogen compounds in field studies (Adeogun et al., 2016b; Cocci et al., 2015; Houtman et al., 2007; Palermo et al., 2008). Increased levels of VTG were found in seabream juveniles following  et al., 2010). In addition, the hepatocyte VTG exposure to PCBs (Calo assay has been suggested as an in vitro screening for identifying estrogen-like substances present in complex mixture (Bjorkblom et al., 2008; Navas and Segner, 2006). In this regard, increased transcript levels of both VTG and ERs were found in trout hepatocytes following exposure to municipal effluents (Gagne et al., 2013). Recently, it has been demonstrated that several estrogenic chemicals (e.g. phthalates and bisphenol A) show the potential to activate the PPAR signaling pathways affecting lipid metabolism in fish hepatocyte cultures (Cocci et al., 2015; Maradonna et al., 2013; Palermo et al., 2016). Adeogun et al. (2016a) have found significantly increased mRNA levels of PPAR isoforms in fish sampled from sites contaminated by PAHs and PCBs. Collectively these findings suggest the potential of PPAR gene expression as endpoint in evaluating the effects of environmental pollutants on metabolic and energetic processes in fish. Among the different biological approaches for evaluating the effects of EDCs in the aquatic environment, the ecotoxicogenomics techniques are considered highly suitable assessment approaches to identify specific pathways of toxicity and to serve as diagnostic tools (Villeneuve et al., 2011). Thus, our objective in this study was to detect endocrine and metabolic effects of EDCs in surface waters of three marine areas in the central Adriatic Sea by means of a toxicogenomic approach using seabream hepatocytes. Primary fish hepatocyte culture represents a validated tool that has been extensively used to measure summated endocrine activities of environmental pollutants by looking at the toxic mode of action. Specifically, changes in gene expression profiles of ERa/VTG, PPARa/Stearoyl-CoA desaturase 1A (SCD1A) and metallothionein (MT) were used to investigate the effect of seawater extracts on both E2 signaling pathway and fatty acid homeostasis. In addition, gene expression patterns were correlated with seawater levels of PCBs and PAHs which have been previously characterized as the contaminants of major concern in the coastal areas of the Adriatic Sea (Quero et al., 2015). In this regard, gene expression of cytochrome P4501A (CYP1A) was employed as sensitive biomarker of environmental exposure to PCBs and PAHs.

2. Materials and methods 2.1. Water samples collection and solid phase extraction (SPE) Data for this study were collected from three locations along the western coast of the central Adriatic Sea (Fig. 1) during 2014 (August through October). All the cruises were conducted aboard of the M/N ECO1. The first location was situated in the coastal area of the Conero's Promontory south of Ancona (A; Marche Region, Ancona-Italy) while the second one (B) was the area immediately inside and outside of the Marine Protected Area (MPA) of Torre del Cerrano (Abruzzo Region, Teramo-Italy). The third sampling area (C) was located in the MPA of the Tremiti Islands (Apulia Region, Foggia-Italy). This location was expected to be the reference site. The data were collected along 3 transects in both designated areas, A and B. Each transect contained 4 sampling stations located at different distance from the coast (0.5- 1.5- 3e5 nm; Fig. 1). Samples were also taken along the coastline of Tremiti Islands (9 sampling stations; SS 1e9). A total of n. 33 CTD (Conductivity Temperature Depth) casts and n. 33 surface water samples were collected and analyzed for nutrient concentrations and for EDC effects by the bioassay. Sample locations were based on the presence of river mouths as a representative fluvial ecosystem located downstream of industrial and agricultural activities areas or in areas characterized by intense maritime traffic. Basic physical properties of the water samples were measured with a SeaBird Electronics SBE 19-plus CTD equipped with a Turner Cyclos fluorometer. The 4 Hz CTD data were processed according to Unesco (UNESCO, 1988) standards, and pressure averaged to 0.5 db intervals. Nutrient water samples were filtered (GF/F Whatman, 0.7 mm) and stored at 22  C in polyethylene vials. Nutrients (ammonium-NH4, nitrite-NO2, nitrate-NO3 and orthophosphatePO4) were analyzed colorimetrically (Parsons et al., 1985). Absorbencies were measured with an AxFlow quAAtro AutoAnalyzers. Dissolved Inorganic Nitrogen (DIN) was calculated as the sum of NH4, NO2, and NO3 concentrations. Surface seawater samples were collected using Niskin bottles (1.7 L) and transferred into labeled polypropylene containers, earlier cleaned with methanol and ultrapure water. Samples were kept at 4  C in the dark and transferred to the laboratory, where they were filtered through a 0.2 mm Nalgene filter (Nalge Nunc International, Rochester, NY, USA) to remove suspended solids and microorganisms. Solid-phase extraction (SPE) was carried out, as described by (Gong et al., 2003). Briefly, all SPE cartridges were sequentially conditioned with 3  5 mL methanol and 2  5 mL ultrapure water. The stored seawater was prewarmed to room temperature, and 1 L water sample was then passed through a 6-mL 500 mg C18 cartridge at a flow rate of 10 mL/min. After all seawater was percolated, each cartridge was flushed with 3  5 mL ultrapure water to remove salts. Cartridges were then centrifuged at 3000 rpm for 10 min, and then a gentle air stream was passed for 10 min to remove most of the water retained in the cartridges. Chemical compounds retained in the cartridges were then eluted with 3  5 mL methanol. The elute was concentrated to a volume of 0.5 mL at 40  C under a reduced pressure of 210 mBar using the Rotavapor R-100 (BUCHI Labortechnik). The concentrated elute was transferred to a 1.5 mL Eppendorf tube and further concentrated to 50 mL (20000 with respect to the original seawater sample) and stored at 4  C in the dark before exposure to the bioassay. 2.2. Determination of PAHs and PCBs For chemical analyses, seawater extracts from the first three sampling stations within each transect (inside both the areas, A and B) were pooled into one sample representing sites within 3 nm of

P. Cocci et al. / Marine Environmental Research 123 (2017) 25e37

27

Fig. 1. Map of the study area. Conero's Promontory south of Ancona (A); Torre del Cerrano Marine Protected Area (B) and Tremiti Islands Marine Protected Area (C). Blue points indicate sampling stations along each transect. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

shore; on the contrary, samples from sites at 5 nm (included as reference stations) were processed individually. Samples from area C were pooled into two samples (SSI and SSII) representing the entire area. The PCB (PCB 28, 52, 101, 118, 138, 153, and 180) concentrations were determined by a gas chromatograph/mass selective detector (GC/MSD) (Hewlett Packard, Palo Alto, CA, USA; HP-6890 with HP 5973) using an appropriate pure standards solution. The PCB mix standard at a concentration of 10 mg L1 in iso-octane was supplied by Dr. Ehrenstorfer (Ausburg, Germany). Standard working solutions at various concentrations were prepared daily by appropriate dilution of the stock solutions with hexane. Hexane for residue analysis was purchased from Riedel-de Haen (Seelze, Germany). Separation was performed on an HP 5 MSI column (30 m  0.25 mm X 0.25 mm film thickness) following a previous published method (Sagratini et al., 2008). An HP Chem workstation was used with the GC/MS system. All injections were splitless and the volume was 1 ml. The flow rate (He) was 0.8 mL min1. The injector temperature was 270  C. The column temperature programme was from 60  C (3 min) to 190  C at 8  C min1, from 190  C (18 min) to 300  C at 15  C min1, then at 300  C for 10 min. Data were acquired in the electron impact (EI) mode (70 eV) using the selected ion monitoring (SIM) mode. SIM ions and time conditions for each PCB congener was reported in Supplementary data, Table 1. Limit of detection (LOD) and limit of quantification (LOQ) were estimated on the basis of 3:1 and 10:1 signal-to-noise ratios obtained with standards containing the compounds of interest at low concentration levels. LODs were in the range 0.045e0.141 ng L1 and the LOQ in the range 0.134e0.423 ng L1 for the 7 PCBs analyzed (Supplementary data, Table 2). PAHs identification and quantification in the seawater extracts were performed using an HPLC system (Ultimate 3000, Thermo

Scientific, Waltham, MA, USA) equipped with a diode array (PDA100) and fluorescence (RF-2000) detectors. The analytical procedure was verified with the external standard calibration. Calibration curves were established using a serial dilution from a standard PAH solution (EPA 610 PAH Mix), purchased from Supelco, Bellafonte, PA, USA. HPLC analysis was performed as described by Filipkowska et al. (2005), with slight modifications. A gradient program on a Hypersil Green PAH analytical reverse-phase column (2.1  150 mm, 1.8 mm, 120 Å) was used. The acetonitrile-water mobile phase gradient had an initial composition of 40% methanol (held for 2 min) that, after 17 min, was increased to 100% over 20 min and then returned to initial conditions. Stop time was set at 40 min. The flow rate was 0.3 mL min1, at temperature of 25  C. LODs were in the range 0.010e0.300 ng L1 and the LOQ in the range 0.030e0.980 ng L1 for the 11 PAHs analyzed (Supplementary data, Table 3). 2.3. Isolation of fish hepatocytes Gilthead seabream (Sparus aurata) hepatocytes were isolated following the method described by Centoducati et al. (2009) with slight modifications (Cocci et al., 2015). Following acclimation, fish were randomly anaesthetized using 3-aminobenzoic acid ethyl ester (MS-222; Sigma; 0.1 g L1) within 5 min after capture, and sacrificed by decapitation. The liver tissue was aseptically harvested to obtain hepatocytes under a laminar flow hood. The detailed procedure for the isolation of seabream hepatocytes was described in our previous publication (Cocci et al., 2015). Following the isolation phases, purified hepatocytes were suspended in Leibovitz L-15 phenol red-free medium supplemented with 10% FBS, antibiotic-antimycotic solution (100 U/ml) and 10 mM HEPES. The cell density was estimated in a counting Burker Chamber and

28

P. Cocci et al. / Marine Environmental Research 123 (2017) 25e37

Table 1 List of primers used in this study. Gene

Primer sequence (50 - 30 )

Genebank

Reference

CYP1A

GCATCAACGACCGCTTCAACGC CCTACAACCTTCTCATCCGACATCTGG CTGGTGCCTTCTCTTTTTGC TGTCTGATGTGGGAGAGCAG CTGCTGAAGAGGGACCAGAC TTGCCTGCAGGATGATGATA GCAGCCTGTGAGTCTTGTGAGTGA CTCCATCAGGTCTCCACACAGC CGGAGGCGGAGGCGTTGGAGAAGAAGAGG GAGACGGCGTACAGGGCACCTATATG CACATGCACAAACTGCTCCT CAGCTGATGTTGCAGACTCC GCATTTATCAGACCCAAAACC AGTTGATAGGGCAGACATTCG

AF011223

Perez-Sanchez et al., 2013

AF136979

Ribecco et al., 2011

AF210428

Cabas et al., 2012

AY590299

Fernandez et al., 2013

JQ277703

Benedito-Palos et al., 2013

X97276

Ribecco et al., 2011

AY993930

Perez-Sanchez et al., 2013

ERa VTG PPARa SCD1A MT 18s rRNA

viability of hepatocytes used for experiments was always over 90%, as assessed with Trypan blue exclusion assay (Sun et al., 2010). Animal manipulation was performed according to the recommendations of the University Ethical Committee, to the European Union directive (2010/63/EU) for animal experiments and under the supervision of the authorized investigators. 2.4. Hepatocytes culture and exposure Cells were seeded on 24-well Falcon Primaria™ culture plates (1  106 cells per well) in Leibovitz L-15 phenol red-free medium supplemented with 10% FBS, antibiotic-antimycotic solution (100 U/ml) and 10 mM HEPES. Cells were cultured for 24 h in an incubator (3% CO2) at 23  C before exposure to allow attachment. Then, L-15 phenol red-free medium culture medium was removed and hepatocytes were exposed to medium containing the vehicle (methanol, MeOH, final concentration 0.01%), seawater extracts (final concentration 10) or E2 (106 M; Anderson et al., 1996; Madigou et al., 2001; Turker and Takemura, 2011). During experimental procedures hepatocytes were incubated in an incubator (3% CO2) at 23  C for 48 h. After 24 h of culture 90% of the medium was removed and replaced with fresh appropriate medium. Six independent wells were set-up for both controls and seawater extracts. After the end of exposure, all cell layers remained attached to the bottom of the plates. At this point, cell viability was again assessed by microscopic examination of the cell morphology and the Trypan blue exclusion test.

reported in Table 1. The reaction included: 12.5 mL 2  qSTAR SYBR Master Mix Kit (OriGene Technologies), 1 mL each of forward and reverse primers (both 10 mmol/L), 0.5 mL cDNA template, and sterile distilled water to a final volume of 20 mL. The expression of individual gene targets was analyzed using the Mx3000P Real-time PCR system (Stratagene, La Jolla, CA, USA). Thermo-cycling for all reactions was for 15 min at 95  C, followed by 40 cycle of 15 s at 95  C and 30 s at 56  C. Fluorescence was monitored at the end of every cycle. Melting curve analysis demonstrated that a single product was generated during the reaction that represents the specific. Results were calculated using the relative 2DDCt method (Livak and Schmittgen, 2001) and expressed as normalized fold expression corrected for 18s rRNA and with respect to control levels. 2.6. Statistical analyses All statistical analyses were performed using R (R Development Core Team, 2011) and JMP (SAS Institute Inc.). q-PCR results were expressed as normalized fold expression corrected for 18s rRNA and with respect to control (vehicle) levels. A one-way analysis of variance (ANOVA) was used to determine the difference among exposure groups, while critical differences between treatments (vehicle control vs water extracts from the different sampling stations) were appraised using the Dunnett's multiple comparisons test. Principal Component Analysis (PCA) and a linear discriminant analysis (LDA) were conducted on the data matrix containing all gene expression values for site classification purposes (Cocci et al., 2015). The results were considered significant when p < 0.05.

2.5. Quantitative realtime PCR (q-PCR) 3. Results and discussion After exposure, medium was carefully removed and cells were lysed by adding the TRIzol® reagent (Invitrogen Life Technologies, Milan, Italy). Total RNA was isolated according to the manufacturer's specifications. DNase digestion (2 U, 30 min, 37  C; Ambion, Austin, TX) was performed to eliminate genomic DNA contamination. RNA concentration and purity were assessed spectrophotometrically at absorbance of 260/280 nm, and the integrity was confirmed by electrophoresis through 1% agarose gels stained with ethidium bromide. The complementary DNA (cDNA) was synthesized from 4 mg of total RNA in 20 mL of total volume reaction using random hexamers (50 ng mL1) and 200 U of SuperScript™ III RT according to manufacturer's instruction (Invitrogen Life Technologies, Milan, Italy). SYBR green-based real-time PCR was used to evaluate expression profiles of CYP1A, ERa, VTG, PPARa, SCD1A and MT. Analysis of the 18s rRNA gene expression confirmed that its expression was unaffected by exposure to seawater extracts or E2, and thus it is considered to be an appropriate reference gene for the qPCR analysis (data not shown). All the primer sequences are

3.1. Physical properties of seawater at the study area Surface distribution of salinity, fluorescence and DIN are showed in Fig. 2. The area A shows a cost-offshore gradient with the salinity increasing and nitrogen decreasing. During autumn rivers runoff along the Italian coast enriching seawater in nutrients and organic matter is more evident than during summer (Socal et al., 2008). The lowest salinity (34.8e35.3) and the highest nitrogen values (6.8e9.6 mmol L1) of the area are detected in the north-western stations. The fluorescence is very low as expected in this period. The area B shows similar pattern in salinity distribution though it is less evident. In particular, the lower salinity values (34.4e34.8) are detected at the coastal stations located in front of the Vomano River's mouth. The area is characterized by low nitrogen that was depleted probably due to uptake from phytoplankton groups as evidenced by relatively high fluorescence values along shore (Eyre and Twigg, 1997; Zavatarelli et al., 1998). The Area C was located at

P. Cocci et al. / Marine Environmental Research 123 (2017) 25e37

29

Fig. 2. Surface distribution of salinity, fluorescence and DIN in the study area.

the open sea and did not appear influenced by freshwater input coming from the Italian rivers that flows southern along shore. The area shows high salinity (37.8e37.9) compared to other ones. The high salinity is associated with low fluorescence and low nitrogen values. All the three areas show very low values of phosphate (0.02e0.04 mmol L1, not shown) as expected in the Adriatic Sea that is a phosphorus limited basin (Zavatarelli et al., 1998).

3.2. PCBs and PAHs seawater concentrations The highest SPCBs concentration was found in samples from area B (transect I), followed by samples from the area A (transects I and II) (Table 2). On the other side, the lowest SPCBs concentration was found along the coastal waters of area C. Since the minimal concentrations of PCBs that exert a negative impact on aquatic

Table 2 Sum of all PCB congeners and PAHs detected in studied areas (A, Marche Region, Ancona-Italy; B, Abruzzo Region, Teramo-Italy; C, Apulia Region, Foggia-Italy). Area

Transect

Sampling site

ƩPCBs ng L1

ƩPAHs (LMW) ng L1

ƩPAHs (HMW) ng L1

ƩPAHs ng L1

A

I

1-3 nm 5 nm 1-3 nm 5 nm 1-3 nm 5 nm

6.260 3.356 5.721 3.609 4.168 3.086

11.412 n.d. 19.598 0.102 19.269 n.d.

19.116 1.863 5.268 1.062 0.611 n.d.

30.528 1.863 24.865 1.164 19.880 n.d.

1-3 nm 5 nm 1-3 nm 5 nm 1-3 nm 5 nm

10.650 2.065 3.344 3.366 3.050 n.a.

80.541 n.d. 14.814 n.d. 30.114 n.a.

18.041 2.505 n.d. n.d. 2.394 n.a.

98.582 2.505 14.814 n.d. 32.508 n.a.

1e4 5e7

1.281 2.673

0.581 1.022

n.d. 0.450

0.581 1.472

II III

B

I II III

C

SSI SSII

n.d.: not detectable; n.a.: not analysed.

30

P. Cocci et al. / Marine Environmental Research 123 (2017) 25e37

organisms, as reported by EPA (2010) and Silva et al. (2009), are 30 ng L1 and 1e2 mg L1, respectively, the overall concentrations detected in our study areas are at an acceptable level. Regarding the congeners that are considered as the markers of environmental contamination by PCBs (Webster et al., 2013), we found detectable levels of only three of them, namely PCB 28, 52 and 138 (Supplementary data, Tables 4e6). PCB 52 and 138 were found at higher concentrations with respect to the other PCBs analyzed (i.e. PCB 28, 101, 118, 153, and 180). The highest levels of PCB 52 were determined in samples from transect I of both the areas, A and B; the lowest in samples collected from the area C. Similarly, the highest level of PCB 138 was found along the transect I located in the area B. Interestingly, samples from area C displayed conspicuous amount of the same molecule (1.010e2.142 ng L1), meanwhile the lowest levels of PCB 138 were found along the transect III in the area A. These findings corroborate previous research documenting the PCB 138 congener is the most prevalent in the Po River Prodelta site (Quero et al., 2015) and along the Abruzzo coast of the Adriatic Sea (Perugini et al., 2004). Analysis of the seawater has also revealed the presence of PAHs (Table 2). Eleven of the sixteen US EPA priority pollutants were determined (Keith, 2015). The total PAHs concentration in water ranged from 0.581 to 98.582 ng L1 with an average of 13.050 ng L1 for the area A, 29.682 ng L1 for the area B, and 1.026 ng L1 for the area C (Supplementary data, Tables 4e6). The results showed a high concentration of PAHs along the first two transects inside the area A and the transects I and III inside the area B (Table 2). Specifically, high PAH levels were found in the first transect near Ancona harbour (area A) and around the mouth of the Vomano River (transect I, area B). The highest concentrations of PAHs were detected in water samples from the area B with the exception of the samples collected along the transect II that is located exactly inside the MPA Torre del Cerrano. Spatial distribution of PAH concentrations perfectly matches the PCBcontaminated sites and are obviously related to river runoffs and marine traffic. Overall, the levels of total PAHs detected in our study area are within the range previously observed in the northern Adriatic Sea (Manodori et al., 2006; Penko, 2010). Considering that the maximum concentration of total PAHs for protection of aquatic life is 200 ng L1 for EU and 30 ng L1 for EPA (EPA, 2009; Nasher et al., 2013; Omayma et al., 2016), the detected PAH levels could be taken as representing contamination of the seawater, especially in area B. The study areas have revealed a PAH molecular distribution dominated by low molecular weight (LMW) PAHs, this is due to their high water solubility and low molecular mass. In particular, the higher concentration of individual PAHs varied from LMW and more volatile Naphthalene (17.204 ng L1), Phenanthrene (31.802 ng L1) and Anthracene (2.142 ng L1) to the high molecular weight Benzo[b]fluoranthene (4.144 ng L1), Benzo[k]fluoranthene (8.896 ng L1), Dibenz[a,h]anthracene (12.398 ng L1), Indeno [1,2,3-cd]pyrene (0.219 ng L1) and Benzo[ghi]perylene (0.648 ng L1). Interestingly, high levels of Acenaphthylene (36.774 ng L1), a LMW PAH normally used to make dyes, plastics and pesticides (HSDB, 2001), were found along the transect I within the area B characterized by activities such as textile manufacturing, paint and chemical productions (CRESA, 2016). In addition, the concentrations of Benzo[a]pyrene (0.180 ng L1) were detected exclusively in seawater samples from the area A (transect I).

Fig. 3. Changes in CYP1A mRNA levels in Sparus aurata hepatocytes exposed to seawater extracts from sampling stations located inside the three investigated areas (A, B and C). Values represent fold change relative to vehicle (Veh) control ± SD of three independent experiments. Asterisks denote a significant (p < 0.05) difference with respect to the Veh. “#” denotes samples not tested.

3.3. Effect of seawater extract samples on CYP1A mRNA levels The observed presence of PAHs and PCBs in seawater from specific transects within the study area led us to monitor the induction of CYP1A expression, a well-established biomarker of aromatic hydrocarbon exposure in fish. An evident pollution-related

inductive response of CYP4501A mRNA levels was observed in hepatocytes treated with extracts from seawater samples collected along transects I and II (Fig. 3) or transects I and III (Fig. 3) inside the area A and B, respectively. On the contrary, there were no statistically significant differences in CYP1A gene expression among

P. Cocci et al. / Marine Environmental Research 123 (2017) 25e37

31

Fig. 4. Changes in ERa and VTG mRNA levels in Sparus aurata hepatocytes exposed to seawater extracts from sampling stations located inside the three investigated areas (A, B and C) or to 106 M 17b-estradiol (E2). Values represent fold change relative to vehicle (Veh) control ± SD of three independent experiments. Asterisks denote a significant (p < 0.05) difference with respect to the Veh. “#” denotes samples not tested.

control (Veh) groups and hepatocytes exposed to seawater extracts collected inside the area B (i.e. transects II) or in the area C (Fig. 3). Thus, the observed CYP1A mRNA response was in good agreement with the spatial variation in PCB and PAH levels. These results are also consistent with previous observations demonstrating correlation between CYP1A expression and the PAH and PCB seawater contents in fish sampled in the Adriatic Sea (Mihailovic et al., 2006). In that study, the authors showed consistent increase in CYP1A

expression in fish collected at sampling sites with low waterborne concentrations (