In vitro effects of PCBs and OH-PCBs on the basal and dexamethasone-modified thyroid hormone metabolism in chicken liver

,

One consequence of rapid technological and industrial development is the pollution of the environment with toxic substances. Some of these chemicals affect the activity of the endocrine system by disrupting the endogenous hormone signalling pathways, and are therefore called 'endocrine disruptors' (EDs) (Swedenborg et al. 2009). Polychlorinated biphenyls (PCBs) and their hydroxylated metabolites (OH-PCBs) belong to the group of EDs. Due to their similar chemical structure, PCB and hydroxylated PCB may mimic or interfere with the action of the endogenous hormones, so these compounds may disturb the functions of the endocrine system. So far, there are known to be at least two different classes of PCBs: dioxin-like PCBs (dl-PCBs, e.g. 2,3′,4,4′,5-pentachlorobiphenyl; PCB118) and non-dioxin-like PCBs (ndl-PCBs, e.g. 2,2′4,4′,5,5′-hexachlorobiphenyl; PCB153). Both dl-PCB and ndl-PCB congeners have been detected in food items such as dairy products, meat, eggs, fish or vegetables, as well as in fruits (Mihtas et al. 2015;Güzel et al. 2020). Moreover, PCB118 and PCB153 have been detected in human plasma (Schettgen et al. 2012), while the presence of their hydroxylated metabolites has been found in urine (Quinete et al. 2016).
The experimental evidence for the effect of PCBs on DIO activity is incomplete and sometimes inconsistent. The effect of these compounds depends on the species and age of the animal, as well as On the other hand, DIO3 has inner ring deiodination (IRD) activity and degrades T 4 to rT 3 , as well as T 3 to 3,3′-T 2 (Fig. 1). Thus, DIO1 and DIO2 are mostly engaged in the synthesis of T 3 , while DIO3 is responsible for TH degradation and inactivation (Van Der Spek et al. 2017). The deiodinases are tissue-specific enzymes: DIO1 is present mostly in the liver, kidneys and muscles; DIO2 occurs in the brain and the liver; while DIO3 has been identified in almost all tissues (Köhrle 2000;Darras & Van Herck 2012;Orozco et al. 2012;Darras et al. 2015).
To fulfil their physiological role as regulators of gene transcription in the target cells, THs must cross the plasma membrane, and this process is mediated by specific TH transmembrane transporters that belong to the families of the monocarboxylate transporters (MCTs), Na-independent organic anion transporting polypeptides (OATPs) and the L-type amino acid transporters (LATs) (Van Der Deure et al. 2010). They are involved not only in the cellular TH uptake, but also in their efflux from the cell (Darras et al. 2015;Bourgeois et al. 2016). T 4 is mainly transported by the monocarboxylate transporter 8 (MCT8) and the organic anion transporting polypeptide 1C1 (OATP1C1), while T 3 is transported by MCT8 and the monocarboxylate transporter 10 (MCT10), and the L-type amino acid transporter 1 (LAT1) is predominantly involved in 3,3′-T 2 transport (Cosmo et al. 2010).
To verify this hypothesis, in vitro experiments were carried out to examine the effects of PCB118 and PCB153 and their hydroxylated metabolites (4-OH-PCB107 and 3-OH-PCB153, respectively) on: (i) T 3 secretion and T 4 to T 3 conversion; (ii) DIO1, DIO2 and DIO3 concentration; and (iii) the expression of the genes encoding a) these three deiodinases, b) the transmembrane transporters of T 4 and T 3 , and c) the TH receptors in control and dexamethasone (DEX)treated chicken liver explants. DEX is a well-known synthetic glucocorticoid analogue that affects the deiodinase mRNA expression and activity in embryonic chicken livers (Van Der Geyten et al. 1999;Van Der Geyten et al. 2001;Verhoelst et al. 2004;Reyns et al. 2005). In addition, PCBs and OH-PCBs, due to their similar chemical structure to thyroid hormones, may also affect deiodination. Therefore, it was decided to investigate the influence of PCBs and OH-PCBs on the iodothyronine metabolisms in both basic and DEX-modified conditions. Moreover, due to the fact that the effect of DEX on the deiodination process has been well described in the literature, this compound served as an internal control to confirm that the technical conditions did not interfere with the course of the experiments.

Animals and experimental procedure
The experiments and animal procedures were accepted by the Second Local Animal Ethics Committee in Krakow, Poland (Resolution No. 20/2022). Hy-Line Brown hens (n = 30; at the age of 24-28 weeks), purchased from a commercial farm, were kept in individual cages with free access to tap water and feed (commercial compound feed for egg laying hens -DJ, from 'Barbara SP. Z O.O., Turza, Poland) in a 14L:10D light regime. The studies were divided into 2 stages: (1) the preliminary stage (the experiments were carried out on 8 hens), which was aimed at selecting an appropriate dose of tested PCBs and OH-PCBs; and (2) the real stage (the experiments were performed on 22 hens), in which the effects of the PCBs and OH-PCBs, alone and in combination with DEX, on: (i) T 3 secretion to the medium and T 4 to T 3 conversion (Exp. 1, n=8 hens); (ii) deiodinase concentrations (Exp. 2, n=8 hens); and (iii) the mRNA expression of deiodinases, TH transporters and receptors (Exp. 3, n=6 hens) in the hen livers were investigated. In both stages, the birds were decapitated in the morning (between 8 and 10 a.m.) to rule out the influence of physical factors on the daily changes the type of compound being tested and the route of administration. For example, experiments carried out on chicken embryos revealed that dl CB77 (3,3′,4,4′-tetrachlorobiphenyl) injected on Day 4 of incubation decreased the plasma TH concentration during the last stages of the chickens' development; however, its influence on DIO1 and DIO3 activity in the liver was limited. In turn, in the cerebellum and telencephalon, a compensatory increase in DIO2 activity with a concomitant decrease in DIO3 activity was observed (Beck et al. 2006). On the other hand, a previous study showed that both mixtures of PCBs, Aroclor 1242 and Aroclor 1254 (mixtures of PCBs containing 12 carbon atoms in the phenyl rings with an average chlorine content of 42 and 54%, respectively), reduced the hepatic DIO1 activity in chicken embryos (Gould et al. 1999). The literature also includes studies on other species, such as fish and mammals, which have confirmed the diversity of the effects of PCBs on TH metabolism. In the Nile tilapia, Aroclor 1254 increased DIO3 activity in the brain, but decreased DIO1 and DIO2 activity in the liver and brain, respectively (Coimbra et al. 2005). An experiment conducted by Kato et al. (2004) on Wistar and Gunn rats showed that PCB99 (2,2′,4,4′,5-pentachlorobiphenyl) decreased the hepatic DIO1 activity, but only in Gunn rats. These observed differences in the effects of PCBs and OH-PCBs on TH metabolism suggest the need for further studies focused on different species, which will allow for looking at the similarities and differences in the regulatory mechanisms in distant taxonomic groups of animals.
Our recent study (Kowalik & Sechman 2022) revealed that PCBs and OH-PCBs may affect TH synthesis and secretion, as well as iodothyronine metabolism in the chicken thyroid gland, by the changing DIO1 and DIO3 mRNA expression and the protein abundances of each DIO. Moreover, even though OH-PCBs are more water soluble than their corresponding parent PCBs, they still retain a high hydrophobicity, which may potentially lead to an accumulation in the liver and adipose tissues. Despite the extensive knowledge concerning the effects of PCBs on many biological processes, there is still scarce data regarding the impact of OH-PCBs, whose concentrations in the environment are currently increasing. The molecular mechanism of action of these compounds and their effect on iodothyronine metabolic processes in the avian liver, which is the main organ of iodothyronine metabolism, has not been fully explored. Therefore, in this study, we tested the hypothesis that PCBs and their OH-PCBs would affect the metabolism of iodothyronines in chicken liver.

T 3 measurement in the medium
The T 3 concentrations in the collected media were determined using T 3 -RIA-CT kits (Cat No. RIA-4534) from DIAsource (Belgium), according to the manufacturer's instructions. To exclude the influence of the incubation medium on the concentration of the measured hormones, the standard curve and controls were reconstituted in the medium. The quality control data of the RIA kit was presented in the previous publication (Kowalik & Sechman 2022).

T 4 to T 3 conversion in liver homogenate
The T 4 to T 3 conversion in the liver homogenates was determined according to the method described by Darras et al. (1991). Briefly, the liver explants were homogenised in an ice-cold sodium phosphate buffer (0.15 M, pH = 6.5) and centrifuged at 2400 g for 10 min at 4°C. To avoid freeze/thaw cycles, immediately after protein isolation, the protein concentration was determined by Bradford's method using a Pierce Detergent Compatible Bradford Assay . The final concentrations of DTT and T 4 in the incubation volume were 2.4 mM and 0.13 µM, respectively. The blank was prepared without the addition of a protein homogenate. After incubation for 1 h at 37°C, the reaction was stopped with 1 ml of ice-cold 0.63% Brij 35 (polyoxyethylene mono-lauryl ether from Serva Electrophoresis GmbH, Germany). The T 3 was measured using T 3 -RIA-CT kits with the following modification: the standards and controls were dissolved in Brij 35 to avoid the influence of this compound on the T 3 determination. The T 4 to T 3 conversion (minus the blank) was expressed per milligram of protein.

Determination of deiodinase concentrations
The deiodinase concentrations in the liver homogenates were determined as described previously (Kowalik & Sechman 2022). Briefly, the tissues were homogenised in liquid nitrogen and suspended in the RIPA Lysis and Extraction Buffer from Applied Biosystems/Thermo Fisher Scientific (Foster City, USA) containing 10 µl/ml of protease inhibitor cocktail from Sigma-Aldrich (St. Louis, MO, in TH secretion and metabolism. From each hen, the liver was isolated, immediately placed on ice, and divided into small and equal explants. In the first part of the in vitro experiments, in order to show the dose-dependent PCB and OH-PCB action on the conversion of T 4 to T 3 , the liver explants were incubated for 24 h in 1 ml Eagle's medium [from the Laboratory of Sera and Vaccines (Lublin, Poland)], which was supplemented with a 0.05% bovine serum albumin (BSA) and 2 μl/ml antibiotic-antimycotic solution (10,000 units penicillin, 10 mg streptomycin and 25 μg amphotericin B/ml) from Sigma-Aldrich (St. Louis, MO, USA) as a control medium, or with the addition of PCB118 (CAS No. 31508-00-6) and PCB153 (CAS No. 35065-27-1) from AccuStandard (New Haven, USA), as well as 4-OH-PCB107 (4-OH-2,3,3',4',5pentachlorobiphenyl, CAS No. 152969-11-4) or 3-OH-PCB153 (3-hydroxy-2,2',4,4',5,5'-hexachlorobiphenyl, CAS No. 54284-55-8) from Wellington Laboratories (Guelph, Canada) at concentrations ranging from 10 -9 M to 10 -5 M (n=8 in each group). In the second stage of the studies, in three consecutive experiments (Exp. 1: n=8; Exp. 2: n=8; and Exp. 3: n=6), the hen liver explants were incubated for 6 h in the following groups: control, in a medium without PCBs or OH-PCBs; and experimental, where the medium was supplemented with: DEX (Cat No. D1756 from Sigma-Aldrich, St. Louis, MO, USA) at a dose 100 nM (10 -7 M, the dose chosen according to Mörk et al. 2016); PCB118, PCB153, 4-OH-PCB107 or 3-OH-PCB153, at a dose 0.5 × 10 -8 M; and combinations of DEX with the above PCBs and OH-PCBs. The doses of the PCBs and OH-PCBs were chosen on the basis of the literature data that showed their levels in avian tissues (Jaspers et al. 2006;Luzardo et al. 2014), as well as the toxicity equivalent factor (TEF) in birds described by Van Den Berg et al. (1998) and the dose-response data obtained from our preliminary studies (see Table 3).
In each experiment, incubation was carried out at 39°C in an atmosphere of 95% air and 5% CO 2 . After the incubation, the liver explants and media were collected and frozen at -80ºC and -20ºC, respe ctively, until T 3 determination in the medium (by the RIA method), protein determination by Bradford's method (Bradford 1976), T 4 to T 3 conversion (according to the method described by Darras et al. 1991), iodothyronine deiodinases (DIO1, DIO2 and DIO3 by the ELISA method) and the mRNA expression (Real-time qPCR method) in the liver tissue of the following genes: i) enzymes involved in the metabolism of T 4 and T 3 : DIO1 (DIO1), DIO2 (DIO2) and DIO3 (DIO3); ii) proteins involved in transmembrane TH transport: OATP1C1 (SLCO1C1), MCT8 naturation for 15 s at 95°C, followed by 1 min of annealing at 60°C. TaqMan probes for the investigated and reference genes were designed and provided by Applied Biosystems (Table 1 and Table 2, respectively). Before proceeding with the determination of the mRNA expression of the tested genes, six reference genes -glyceraldehyde 3-phosphate dehydrogenase (GAPDH), hypoxanthine-guanine phosphoribosyltransferase (HPRT1), 60S ribosomal protein L13 (RPL13), succinate dehydrogenase complex flavoprotein subunit A (SDHA), TATA-binding protein (TBP) and vimentin (VIM) -were tested to select the most appropriate one. According to the statistical analysis, which was performed with the NormFinder programme, the most stable gene was SDHA (stability value 0.028); while the stability value for the best combination of two genes, VIM and SDHA, was 0.032. For this reason, the SDHA gene was used as a reference gene. Because the reaction efficiency for the tested genes was in the range of 90-110%, the normalised relative quantities (NRQ) of the genes were calculated according to the Pfaffl (2001) method.

Statistical analysis
The data was statistically analysed by a one-way ANOVA, followed by a Tukey's multiple range test (for independent trials) performed using Sigma Stat 2.03 (Systat Software GmbH, Germany). In order to standardise the statistical analysis of the results obtained, and due to the fact that parametric tests are considered stronger compared to non-parametric tests, log transformations were made as needed to maintain the normal distribution of the data (a loga-USA), then sonicated and centrifuged at 12,000 rpm for 20 min at 4°C. The protein was determined by Bradford´s method as described above. The isolated protein was used to measure the concentration of iodothyronine deiodinases in the liver tissue by using chicken DIO1 (Cat No. E12262c), DIO2 (Cat No. 0578c) and DIO3 (Cat No. E12263c) ELISA kits (Wuhan EIAab Science Co. Ltd., China), according to the manufacturer's instructions. The quality control data of the ELISA kits was shown in the previous publication (Kowalik & Sechman 2022).

Gene expression analysis
The total RNA from the liver explants was isolated according to the modified method of Chomczynski & Sacchi (2006) using the TRI reagent from Thermo Scientific (Rockford, IL, USA). The A260/A280 ratios and an electrophoretic separation were performed to evaluate the quality of the extracted RNA. The reverse transcription (RT) and real-time PCR were carried out according to our recent (Kowalik & Sechman 2022) and previous work Katarzyńska-Banasik et al. (2017). Briefly, the RT reaction was performed on one µg of total RNA using the Rever-tAid RT Transcription Kit (from Thermo Scientific) in the Mastercycler personal thermocycler (Eppendorf, Germany) with the following thermal profile: 5 min at 25°C, 60 min at 42°C and 5 min at 70°C. Next, 10-fold diluted cDNA was used in a real-time polymerase chain reaction (qPCR) for the tested and reference genes in a 96-well thermocycler (StepOne Plus, Applied Biosystems, Foster City, CA, USA), according to the following cycling programme: 2 min at 50°C, 10 min at 95°C and 40 cycles of de- 10 -9 to 10 -7 M) elevated the T 4 to T 3 conversion by an average of 21% (p < 0.05) and 46% (p < 0.01), respectively. PCB153 at a concentration of 10 -9 and 10 -8 M augmented the measured conversion by an average of 52% (p < 0.01). All the applied doses of 3-OH-PCB increased the T 4 to T 3 conversion by an average of 50% (p < 0.05-0.01).

Effect of PCBs and OH-PCBs on T 3 secretion and T 4 to T 3 conversion
In the control conditions, the explants of the chicken liver secreted 1.24 ± 0.06 ( Fig. 2A) and 1.18 ± 0.08 ng of T 3 /mg protein (Fig. 2B). In both trials, the DEX inhibited T 3 secretion from the liver explants by 27% and 24%, respectively (p < 0.05; Figs 2A, 2B). PCB118, as well as PCB153, decreased the T 3 concentration in the medium by 28% (p < 0.05; Fig. 2A) and 40% (p < 0.01; Fig. 2B), respectively. On the other hand, 3-OH-PCB153 increased the T 3 secre-rithmic transformation of the data was used only for hypothesis testing). Values are shown as the mean ± SEM (presented on the original scale) and are considered to be significantly different at p < 0.05. The figures were prepared using Grapher 17 (Golden Software Inc., USA).

Selection of the dose of the tested PCBs and OH-PCBs
In the control groups, the conversion of T 4 to T 3 in the liver explants ranged from 152.38 ± 6.27 to 156.53 ± 8.85 pg T 3 /mg protein (Table 3). Depending on the applied dose, all the tested compounds increased the T 4 to T 3 conversion. PCB118 (at doses of 10 -8 and 10 -7 M) and 4-OH-PCB107 (at doses from  Table 3 Effect of PCB118 and PCB153 and their hydroxylated metabolites (4-OH-PCB107 and 3-OH-PCB153) at doses from 10 -9 M to 10 -5 M on T 4 to T 3 conversion in the chicken liver after 24 h of incubation. The T 4 to T 3 conversion was determined according to the method described by Darras et al. 1991. Values are expressed as mean ± SEM (n = 8 per each group) in pg T 3 /mg protein. creased the T 4 to T 3 conversion by 36% and 100%, respectively (p < 0.01; Fig. 2C). PCB153 did not affect the T 4 to T 3 conversion; however, 3-OH-PCB153 elevated this conversion by 43% (p < 0.01; Fig. 2D). Furthermore, all the tested PCBs and OH-PCBs counteracted the inhibitory effect of DEX on T 4 to T 3 deiodination in the chicken liver homogenates (p < 0.01 in comparison to DEX; Figs 2C, 2D). On the other hand, the stimulatory effects of 4-OH-PCB107 and 3-OH-PCB153 on the T 4 to T 3 conversion were reduced by the DEX by 22% (p < 0.05; Fig. 2C) and 34%, respectively (p < 0.01; Fig. 2D).
Effect of PCBs and OH-PCBs on the mRNA expression of iodothyronine deiodinases In comparison with the control group, DEX decreased the DIO1 mRNA level in the chicken liv-tion by 24% (p < 0.05; Fig. 2B). Neither PCB nor OH-PCB influenced the DEX-inhibited T 3 secretion from the liver explants (p < 0.05 ; Figs 2A, 2B). However, this steroid reversed the stimulatory effect of 3-OH-PCB153; as a result, the T 3 secretion in this group was 47% lower in comparison with the control group (p < 0.01; Fig. 2B).
The effects of dl-and ndl-PCB and their OH-PCB congeners on the T 4 to T 3 conversion rate in the liver explants are shown in Fig. 2C and Fig. 2D. In the two control groups, the concentration T 3 (representing the T 4 to T 3 conversion in the liver homogenates) was 125.24 ± 9.22 and 122.14 ± 9.39 pg T 3 /mg protein, respectively (Figss 2C, 2D). In both trials, DEX inhibited the T 4 to T 3 deiodination by 44% and 48%, respectively (p < 0.01; Figs. 2C, 2D). In comparison to the control group, PCB118 and 4-OH-PCB107 in- to T 3 conversion in the liver homogenates were determined by radioimmunoassay (RIA) and according to the method described by Darras et al. 1991, respectively. Values are expressed as the mean ± SEM (n = 8 per each group). Different letters denote statistically significant differences (p < 0.05).
with the DEX-treated group, in the PCB118 + DEXtreated group, the DIO1 mRNA level was lower by 21% (p < 0.05; Fig. 3A). There were no significant alterations between the treated groups with respect to ndl-PCBs (Fig. 3B).
The mRNA expression of the OATP1C1 gene was not detected in the livers of the domestic hens.

Effect of PCBs and OH-PCBs on the mRNA expression of the receptors involved in target gene expression
The effects of the tested PCBs and OH-PCBs on the TRα and TRβ0 mRNA expression are shown in Fig. 6. Only 4-OH-PCB107 decreased the mRNA expression of TRβ0 by 22% (p < 0.05; Fig. 6C). In the other groups, no significant changes in the mRNA expression of TRα and TRβ0 were observed (Fig. 6).

Discussion
Although the production of PCBs was banned in the year 2001 by the Stockholm Convention, PCBs and OH-PCBs, like other xenobiotics, are still present in the environment and therefore pose a threat to the health of living organisms (Borja et al. 2005;Pocar et al. 2006;Darras 2008). PCBs and OH-PCBs are detected not only in the environment (Borja et al. 2005;El-Rahman et al. 2019;Fang et al. 2020;Aziza et al. 2021) and in food products (Arnich et al. 2009;Mihats et al. 2015), but also in the blood plasma and urine of humans and animals (Schettgen et al. 2012;Koh et al. 2016;Quinete et al. 2016). Furthermore, the OH-PCBs produced by the oxidation of parental PCBs through a variety of mechanisms (such as metabolic transformations in living organisms or abiotic reactions with hydroxyl radicals) are still highly toxic, and probably play an important role in the resistance of PCBs to biodegradation (Kawano et al. 2005;Tehrani & Van Aken 2014). Both PCBs and their more water-soluble hydroxyl metabolites can easily accumulate in the adipose tissue and liver, or bind to the plasma proteins, because they retain a high hydrophobicity (Tampal et al. 2002;Tehrani & Van Aken 2014).

Effect of PCBs and OH-PCBs on the mRNA expression of transmembrane TH transporters
In comparison with the control group, DEX did not affect the mRNA expression of the MCT8 (Figs 5A, 5B) and MCT10 (Figs 5C, 5D) transporters in the chicken explants, but increased the LAT1 mRNA levels by an average of 42% (p < 0.01; Figs 5E, 5F).
All the tested PCBs and OH-PCBs, alone and in combination with DEX, did not influence the expression of MCT8 mRNA (Figs 5A, 5B). The basal MCT10 mRNA expression was inhibited in the groups incubated in a medium supplemented with PCB118 and 4-OH-PCB107 by 19% and 32%, respectively (p < 0.05; Fig. 5C). However, the addition of DEX to the incubation medium eliminated these inhibitory effects (Fig. 5C). PCB153 and 3-OH-PCB153 alone had no influence on the MCT10 the effects of two different PCB classes, dl-PCBs and ndl-PCBs, represented by PCB118 and PCB153, respectively, and of their hydroxylated metabolites (4-OH-PCB107 and 3-OH-PCB153) on the T 3 secretion from liver explants, T 4 to T 3 conversion, iodothyronine deiodinase concentration, and the mRNA expression of the main genes involved in the metabolism, transport and THs receptors in chicken livers.
Even though there is data in the literature on the influence of PCBs on the THs metabolism, it is difficult to clearly define the direction of the influence of these compounds. Moreover, their molecular mechanism of action in the cell, as well as the influence of OH-PCB on the metabolism and transport of THs in chicken liver, have not been fully explained. Therefore, this study was undertaken to investigate man et al. 2016) were also taken into account. Based on the abovementioned literature data, for both PCB and OH-PCB, doses from 10 -9 M to 10 -5 M were selected and used in the preliminary experiment. The parental PCBs increased the T 4 to T 3 conversion in the groups incubated with the addition of PCB118 at concentrations of 10 -8 M and 10 -7 M, and PCB153 at concentrations of 10 -9 M and 10 -8 M. The same stimulating effect was observed in the groups incubated with 4-OH-PCB107 at concentrations ranging from 10 -9 M to 10 -7 M, as well as with 3-OH-PCB153 at all the tested concentrations. The observed effect depended on the concentration and form of the tested xenobiotics; while T 4 to T 3 deiodination was stimulated more by the OH-PCBs in comparison to the parental PCBs.
Because changes were observed mainly in the groups incubated in a medium supplemented with the tested PCBs and OH-PCBs at concentrations of 10 -8 M and 10 -7 M, in order to standardise the plan and layout of the further experiments, an intermediate dose of 0.5 x 10 -8 M was chosen and applied in this study.
The first, preliminary stage of our research was to establish the concentrations of PCB118 and PCB153, as well as their hydroxylated metabolites, which would be used in further in vitro experiments. Jaspers et al. (2006) analysed the concentrations of several parental PCBs, including PCB118 and PCB153, in the livers of various aquatic and raptor birds. They showed that the median concentrations of PCB118 and PCB153 (ng/g fat mass) in this organ were respectively: 890 and 17,000 in the grey heron (Ardea cinerea), or 610 and 2,300 in the great crested grebe (Podiceps cristatus). In turn, in the representatives of raptors, the median concentrations of PCB118 and PCB153 (ng/g fat mass) were: 1,800 and 45,000 in the barn owl (Tytoninae), 71 and 2,000 in the common buzzard (Buteo buteo) or 6,000 and 31,000 in the sparrowhawk (Accipiter nisus). For the selection of the appropriate doses, values of the TEF coefficients of the applied PCBs evaluated for birds (Van Den Berg et al. 1998)   confirmed the presence of all three types of iodothyronine deiodinases at the mRNA and protein levels in the chicken livers. The determination of these deiodinases by the chicken-specific ELISA method revealed that their concentrations in chicken liver are arranged in the following order: DIO1>DIO3>DIO2.
A similar relationship was found in humans who were tested for the activity of these three types of deiodinases (Richard et al. 1998). In our research, DEX diminished DIO1 and DIO3, and did not influence the DIO2 mRNA expression and protein concentration in the liver explants. These results also confirm our earlier hypothesis (based on the obtained results and literature data; Darras et al. 1996;Van Der Geyten et al. 1999;Van Der Geyten et al. 2001;Reyns et al. 2005) that the effect of DEX on iodothyronine metabolism catalysed by deiodinase (T 3 concentrations, as well as the protein and mRNA level of deiodinase) depends on the age of the chicken. The data in the literature shows a different influence of DEX on the metabolism of THs and deiodinase activities in embryonic and post-hatch chickens: in 18-day-old embryos, a single injection of DEX did not affect DIO1 activity in the liver, but it rapidly (within just 2 h after the injection) decreased the DIO3 activity (Reyns et al. 2005). These changes were accompanied by a decrease in the DIO3 mRNA expression, but without an influence on the DIO1 mRNA levels following a DEX treatment (Van Der Geyten et al. 1999). Nevertheless, it has been found that in 18-day-old chicken embryos, the longacting DEX may increase the DIO1 activity at 24 and 48 h after an injection (Darras et al. 1996).
Alternatively, in 8-day-old chickens after hatching, this glucocorticoid decreases DIO1 but has no effect on the DIO3 activity (Darras et al. 1996). In the literature, there is no data on the effect of DEX on the DIO2 mRNA levels in chicken liver; however, in all brain parts of 18-day-old embryonic chickens, DEX has been shown to affect the T 3 production capacity by an upregulation of DIO2 activity (Reyns et al. 2005).
Our experiment revealed that, in laying hens, PCB118 and 4-OH-PCB107 augmented the DIO1 mRNA level. Moreover, 3-OH-PCB153 increased the DIO2 while 4-OH-PCB107 decreased the DIO3 mRNA expression. On the other hand, in the groups incubated in the medium with the addition of the tested dl-PCBs and their hydroxylated metabolites together with DEX, a greater effect of these compounds was observed than in the groups treated only with PCBs or OH-PCBs. In particular, it seems that dl-PCBs can interact with and/or counteract the effects induced by DEX on the deiodinase mRNA The results of the in vitro experiments indicated that DEX has different effects on T 3 secretion and T 4 to T 3 conversion in the livers of adult chickens, in comparison with embryonic and post-hatch chickens. In our experiment, we found that DEX reduced both the T 3 secretion and T 4 to T 3 conversion in liver explants of the laying hen. Experiments carried out on 18-day-old chicken embryos clearly showed that during their embryonic development, DEX increases the concentration of T 3 in the blood plasma and liver (Darras et al. 1996;Van Der Geyten et al. 1999;Van Der Geyten et al. 2001;Reyns et al. 2005). On the other hand, in 8-day-old post-hatch chicks, DEX tended to lower the plasma T 3 (Darras et al. 1996). These results support the hypothesis that the effect of DEX on T 3 secretion from the chicken liver depends, to a large extent, on the age of the animal.
Our results revealed that the tested PCBs and OH-PCBs affected the basal and DEX-inhibited T 3 secretion from the chicken liver explants. PCB118 decreased the T 3 secretion; while its hydroxylated metabolite, i.e. 4-OH-PCB107, had no effect on the secretion of this hormone. PCB153 and 3-OH-PCB153 also changed the T 3 secretion from the liver explants, but in an opposite manner, i.e. the parental PCB decreased, while the hydroxylated metabolite increased the secretion of this iodothyronine. It is interesting that both PCB118 and PCB153, as well as their hydroxylated metabolites, did not change the DEX-inhibited T 3 secretion. To our knowledge, this is the first in vitro study to examine the T 3 secretion from liver explants. These results suggest that: (i) both parental PCBs, dl-and ndl-PCBs, inhibit T 3 secretion; (ii) the OH-PCB effects are OH-PCB congener dependent; and (iii) the tested PCBs did not influence the DEX-inhibited secretion of T 3 from the liver explants. One of the factors that could explain these changes in the T 3 secretion from the liver explants following PCBs and OH-PCBs treatment is T 4 to T 3 conversion. Our results showed that the investigated PCBs and OH-PCBs stimulated both basic and DEX-inhibited T 4 to T 3 conversion. Moreover, all the tested compounds abolished the inhibitory effect of DEX on the T 4 to T 3 conversion. The observed discrepancies between T 3 secretion and T 4 to T 3 conversion are difficult to explain; however, it cannot be excluded that PCBs and OH-PCBs may reduce the permeability of hepatocyte cell membranes and/or affect the functions of the TH transporters. Nonetheless, this hypothesis needs further investigation.
To explain the observed effect of the investigated PCBs and OH-PCBs on T 3 secretion and T 4 deiodination, we tested the mRNA expression and protein level of iodothyronine deiodinases. This research geois et al. 2016). Our studies showed a lack of OATP1C1 mRNA expression in the chicken livers. Furthermore, the investigated PCBs, OH-PCBs and DEX did not influence the MCT8 mRNA expression and had a slight effect on the MCT10 mRNA expression, because only PCB118 and 4-OH-PCB107 diminished the mRNA level of this gene. These results suggest that the tested compounds hardly affect, or have little effect on T 4 and T 3 transport across the liver cell membrane by the MCT8 and MCT10 proteins. However, it cannot be ruled out that the observed reduction in T 3 secretion following the PCB118 and 4-OH-PCB107 treatment was associated with the decrease in the MCT10 mRNA expression. On the other hand, a change in the LAT1 mRNA level in all the experimental groups suggests that DEX, the tested PCBs and OH-PCBs, as well as the combination of these compounds, may mainly stimulate 3,3′-T 2 transport in hepatocytes. Therefore, the observed increase in the LAT1 mRNA level seems to be related to the enlargement of the T 3 to 3,3′-T 2 deiodination catalysed by DIO1 and DIO3 to protect the cells from too high a concentration of T 3 . Our results are consistent with the studies of Kato et al. (2017), who revealed that rats treated with KC500 (a mixture of seven different PCB congeners) increased the LAT1 mRNA and OATP2 expression in the liver, but these effects were observed at 48 and 96 h for LAT1, and at 96 h for OATP2.
Because PCBs and OH-PCBs have a similar chemical structure to TH, it seems likely that they may bind to TR. To explain the molecular mechanism of the PCB congener, as well as the action of OH-PCBs in the liver, we determined the mRNA expression of the genes encoding two isoforms of TRs, i.e. TRα1 and TRβ0. Only 4-OH-PCB diminished the mRNA expression of TRβ0, which suggests that the tested PCB and OH-PCBs do not directly affect the THdependent gene expression in liver cells, but through changes in the secretion or deiodination of THs, they may change the functioning of hepatocytes. A similar effect was observed in rats where the authors investigated the effect of, inter alia, various PCB and OH-PCB congeners on the ability to bind to TRs in the hepatic nuclei, and the conclusion was that these compounds do not appear to bind to TRs in a competitive manner (Gauger et al. 2004). Cheek et al. (1999) showed that OH-PCBs have a relatively low affinity for the human thyroid receptor in vitro, and furthermore, none of the dl-PCB, PCB 77 and PCB 126 tested bound to TR. However, our results and those of other authors suggest that PCBs and OH-PCBs do not directly compete with TH for TR binding.
levels. Compared to the mRNA expression analysis, the tested compounds showed a greater effect on the concentration of deiodinases in the chicken livers. Similarly to the mRNA expression, PCB118 and 4-OH-PCB107 augmented the DIO1 and did not change the DIO2 and DIO3 concentrations. Moreover, both of these compounds increased the DEX-inhibited level of DIO1 and DIO3. On the other hand, PCB153 and its hydroxylated metabolite influenced the concentrations of all three types of iodothyronine deiodinases, which suggests that ndl-PCB may disturb TH metabolism in the liver. Discrepancies observed between the mRNA expression and the protein levels of iodothyronine deiodinases in the liver explants may be related to differential transcriptional, posttranscriptional and/ or translational regulation. The changes in the T 4 to T 3 conversion, as well as the deiodinase mRNA expression and concentrations demonstrated in our research are correlated. The enlarged T 4 deiodination after dl-PCB treatment may be a result of increased DIO1 mRNA and protein levels. It cannot be excluded that in the group incubated with PCB153, due to the increased concentration of DIO1 and DIO3, T 4 was converted to rT 3 to a greater extent than to T 3 . This possibility is also confirmed by the decrease in the DIO2 level, which may suggest that DIO1 and DIO2 (responsible for the T 4 to T 3 conversion) can counteract each other, and as a result, the deiodination in this direction becomes less effective. Moreover, 3-OH-PCB153, by increasing the concentrations of DIO1, DIO2 and DIO3, may stimulate the deiodination process in each direction. The research on the influence of PCBs on deiodinases mainly concerns the activity of these enzymes. It has been shown that the hepatic activity of DIO1 remained unchanged after PCB77 treatment in chicken embryos (Beck et al. 2006). The same lack of effect was observed in chickens after hatching due to PCB54 (2,2′,6,6′-tetrachlorobiphenyl), PCB77 or PCB80 (3,3′,5,5′-tetrachlorobiphenyl; Gould et al. 1999;Beck et al. 2006). Experiments on rats showed that Aroclor 1254, a mixture of various PCB congeners, reduced DIO1 activity in the liver (Hood & Klaassen 2000). On the other hand, PCB77 and Aroclor 1254 exerted stimulatory effects on the DIO2 activity in the brains of chicken embryos (Beck et al. 2006) and in rat foetuses (Morse et al. 1996). Nevertheless, PCB77 did not affect the hepatic DIO3 activity in chicken embryos (Beck et al. 2006), while the 21-day administration of Arclor1254 increased DIO3 activity in the Nile tilapia (Coimbra et al. 2005).
The transport of THs across the cell membrane requires the presence of special carrier proteins (Bour- Although in vitro studies allow for more a precise control of the experimental conditions, as well as for examining the effects of compounds directly on the tissue, which allows researchers to exclude other mechanisms responsible for eliminating the effects of toxins in the body (e.g. it cannot be ruled out that the observed changes in the liver could be compensated by the xenobiotic metabolism processes taking place in other tissues), it would be very desirable to also conduct in vivo tests, in order to confirm the observed effects of PCBs and OH-PCBs on the iodothyronine metabolism in a chicken's liver.

Conclusions
In conclusion, this is the first in vitro study to describe the effects of PCBs and their hydroxylated metabolites on: basal and DEX-affected T 3 secretion; T 4 to T 3 conversion; iodothyronine deiodinase concentrations; and the mRNA expression of the genes involved in TH metabolism, membrane transport and TRs in chicken liver. Our results suggest that both dl-PCB and ndl-PCB and their hydroxylated metabolites mainly affect the T 4 to T 3 conversion and, to a lesser extent, the T 3 secretion from liver explants. Moreover, PCBs and OH-PCBs may influence the TH metabolism in chicken livers by a modification of the transcription and protein abundance of iodothyronine deiodinases. The changes we observed in the mRNA expression of the MCT8 and LAT1 indicated that PCBs and OH-PCBs may affect TH transport across the hepatocyte membrane. Our results suggest that not only parental PCBs, but also their hydroxylated metabolites may affect liver functions associated with the metabolism of iodothyronines. However, since in vitro studies take place in controlled experimental conditions, in vivo studies are recommended to confirm the observed effects of PCBs and OH-PCBs on iodothyronine metabolism in the liver of the chicken.

Author contributions
Research concept and design: K.K..; Collection and/or assembly of data: K.K., K.