Comparative toxicity of trophic exposures to polybrominated compounds (2, 4, 6 - Tribromophenol and BDE-209) to Oreochromis niloticus (Linnaeus, 1758)


 In the last decades, there has been an increase in demand for new polymers, including flame-retardants compounds, to meet fire prevention by international safety standards. The decabromodiphenyl ether (BDE-209) is a polybrominated diphenyl ether (PBDE) congener still in use worldwide. This compound presents lipophilic properties and so is easily bioaccumulated in the food chain. The 2, 4, 6 - Tribromophenol (TBP) is a PBDE metabolite also used as pesticide and flame-retardant for wood conservation. In the current study, the acute and chronic toxicity of BDE-209 and TBP was evaluated in Oreochromis niloticus through the analyses of redox unbalance, neurotoxicity and histopathological biomarkers after acute (24, 48, 72 and 96 h) and chronic (90 days) exposures to 0.5, 5 and 50 ng/g of the PBDEs. The results showed effects in GST activity and damage to biomolecules in both acute and chronic exposures. Histopathological findings were observed in the acute experiment, while hepatocyte lesions were found in both experiments. Only BDE-209 presented neurotoxic effects. The current study revealed new endpoints related with polybrominated compounds in fish, highlighting the needs to review the risk of exposure to biota and human populations.


Introduction
Polybrominated are persistent and bioaccumulative compounds with worldwide distribution in the environment and biota. In the last few decades, these molecules have been used by industries as wood preservatives, fungicides and ame retardants. The production of 2, 4, 6 -tribromophenol (TBP) exceeded 170,000 tons in 2004 (ECB, 2006), but according to Covaci et al. (2011) only the USA production was estimated between 4,500 to 23,000 tons/year. The use of these compounds as a re retardant in textiles, polyurethanes, plastics, epoxy resins and paper manufacturing, as well as additives or intermediates for the yield of other ame retardants (Weil and Leuchite, 2004) has generated considerable concerns about their toxic effects to humans and wildlife. The consequences are the presence of TBP in surface water (Reinke, 2006), marine sh (Chung et al, 2003), human milk (Ohta et al, 2004) and even human placental levels (Leonetti et al., 2016).
The production of BDE-209 begun in the 1960s, and since 1970s it has been used as ame retardant in a diversity of products, mainly in electrical and electronic industries (English et al., 2016). According to the same authors, the presence of BDE-209 in plastic used in TVs is about 10% and in textiles reached 25% (5 g/m 2 ). This compound is the most dominant PBDE congener in indoor dust (English et al., 2016) leading to an infant ingestion about 55-200 mg/day. High concentration of BDE-209 was detected in waste water and sewage sludge in China (Peng et al., 2009) or even in remote locations, such as the Artic (Tomy et al., 2009). According to Ross et al. (2009), there is a large reservoir of BDE-209 in the environment and the indoor environments are the most important sites of human contamination. Studies also estimate an ingestion of PBDEs via sh consumption (4.4 to 14 ng/kg) in Hong Kong residents (Wang et al., 2011). According to Wang et al. (2011), PBDEs was found in human breast milk, serum and blood (Sjödin et al., 2008), while Leonetti et al. (2016) detected BDE-209 in 50% of placentas from women. More recently, Agnar-Alemany et al. (2017) reported the presence of PBDEs in 90.5% of the sea food samples, particularly in mussels from the Mediterranean Sea.
The aim of the current study was to investigate the acute and chronic trophic effects of TBP and BDE-209 in Oreochomis niloticus through biochemical, histopathological and ultrastructure biomarkers, increasing the knowledge about the risk of exposure to biota or even human populations.
For acute experiments, the tissues were sampling 24, 48, 72 and 96 h after gavage, while for chronic exposure the sampling occurred ten days after the last gavage (80th day). The animals were anesthetized with MS-222 (0.02% in water), weighed and measured for biometric data. After blood sampling from caudal vein, the animals were euthanized by medullar section, and biological samples were stored at -80°C (liver, muscle and brain) or preserved for histopathological (light and electron microscopy) analysis (liver).

Biomarker analyses
Biometric parameters: the hepatosomatic index (HSI) and gonadosomatic index (GSI) were calculated as a relation between liver or gonads weight and the total body weight: HSI or GSI = (OW/TW) x100, where 'OW' means the organ weight and 'TW' means the total body weight. The Fulton's condition factor (K) considers the isometric weight-length relationship and was obtained by the equation: K = (TW/TL 3 ) x100, where 'TW' means total weight and 'TL' the total length.
A standard curve of bovine serum albumin was used to calculate protein concentration (Bradford, 1976). Protein concentration was adjusted to 1 mg mL − 1 for glutathione S-transferase and acetylcholinesterase, and 2 mg mL − 1 for protein carbonylation assays. Original protocols were adjusted for use of 96-well microplates and all readings were performed in a Varioskan LUX Multimode Microplate Reader (Thermo Scienti c) Glutathione S-transferase (GST) activity: 20 µL of samples (liver) and 180 µL of the reaction medium (1.5 mM glutathione (GSH), 2.0 mM 1-chloro-2,4dinitrobenzene (CDNB) in 0.1 M phosphate buffer, pH 6.5) were added to a 96-well microplate, and the absorbance increased was immediately measured for 5 min at 340 nm. The molar extinction coe cient for CDNB (9.6 mM − 1 cm − 1 ) was used to calculate GST activity (Keen et al. 1976).
Non-protein thiols (NPT): 200 µL of samples (liver) and 50 µL of trichloroacetic acid (10%, Sigma) were mixed in a 1.5 mL tube and centrifuged at 1,000 g for 15 min at 4°C. Then, 50 µL of the supernatant, 230 µL of tris-base buffer (0.4 M, pH 8.9) and 20 µL of 2.5 mM DTNB (in 25% methanol) were added to a 96well microplate. After incubation for 5 min at room temperature, the absorbance was measured at 415 nm. A standard curve of GSH was used to calculate NPT concentration (Sedlak and Lindsay, 1968).
Lipid peroxidation (LPO): 200 µL of samples (liver) and 800 µL of reaction medium (0.1 mM xylenol orange (Acros Organics), 25 mM H 2 SO 4 , 4.0 mM hydroxytoluene butylated (Acros Organics) and 0.25 mM ferrous ammonium sulfate, prepared in this order in 90% methanol) were mixed in a 2 mL tube. After incubation for 20 min at room temperature, the tubes were centrifuged at 9,000 g for 10 min and 200 µL of supernatant were added to a 96-well microplate.
The absorbance was measured at 570 nm and the molar extinction coe cient for hydrogen peroxide and butyl hydroperoxide (4.3x10 4 M − 1 cm − 1 ) was used to calculate the lipid hydroperoxides concentration (Jiang et al. 1991(Jiang et al. , 1992. Protein carbonylation (PCO): samples (liver and brain) were distributed into two 1.5 mL tubes (blank and test) and mixed with either 500 µL of 10 mM 2,4dinitrophenylhydrazine in 2 M HCl (test tubes) or with 2 M HCl (blank tubes). After incubation for 1.5 h at 37°C, 1 mL of 28% trichloroacetic acid was added and the tubes were centrifuged (9,000 g, 10 min, 4°C). The protein pellet was resuspended with 1 mL of ethyl acetate/ethanol (1:1), mixed and centrifuged (9,000 g, 10 min, 4°C). This procedure was repeated three times until supernatant was clear. Then, the pellet was resuspended in 500 µL of 6 M guanidine hydrochloride (Sigma-Aldrich), mixed and centrifuged at 9,000 g for 3 min at 4°C. Supernatants (200 µL) were added to a 96-well microplate and the absorbance was measured at 360 nm. The molar extinction coe cient for hydrazones (2.1 x 10 4 M − 1 cm − 1 ) was used to calculate the carbonyl concentration of samples (Levine et al. 1994).

Histopathological procedures
Light microscopy: Liver samples were xed in ALFAC (70% ethanol, 4% formaldehyde and 5% acetic acid) for 16 h and washed with 70% ethanol solution, dehydrated in a crescent ethanol series, diaphanized in xylol and included with Paraplast resin (Sigma®) in a Tissue Processor (Micron STP120TM) and resin dispenser (Micron STP120TM), according to Oliveira Ribeiro et al. (2012). 5 µm-thick cross sections were obtained using a semi-automatic microtome Leica TM and stained with Hematoxylin and Eosin. The histopathological study was performed according to Bernet et al. (1999) lesion index and the images registered by an Olympus BX51 Microscope from the Center for Advanced Fluorescence Technologies at Federal University of Paraná (CTAF-UFPR). The lesion index was calculated using the following formula: IL = Σalt (a x W), where: IL = lesion index, Σalt = sum of damages/changes, a = degree of occurrence and W = factor of importance.

Transmission Electron Microscopy
Liver samples were xed at room temperature (2.5% glutaraldehyde, 4% paraformaldehyde, 0.05 M CaCl 2 in 0.1 M sodium cacodylate, pH 7.2-7.4) for 2 h, washed with 0.1 M sodium cacodylate buffer, post-xed in 1% osmium tetroxide (Sigma-Aldrich®) in 0.1 M cacodylate buffer for 1 h, washed in 0.1 M cacodylate and dehydrated in graded ethanol series (Merck®) and propylene oxide (EMS®). The embedding was performed in PoliEmbed 812 DER736 (Electron Microscopy Science®) to obtain ultra ne sections (70 nm thick) using a Leica Ultramicrotome. The sections were contrasted by uranyl acetated (20 min) and lead citrate (5 min) according to Oliveira Ribeiro et al. (2012), and observed in a JEOL TEM 1200 EXII from Electron Microscopy Center at Federal University of Paraná.

Statistical procedures
The dependent variables were previously checked for the assumptions of residual normality and homoscedasticity using the Shapiro-Wilk and Levene test, respectively. The analysis parameters were submitted to the three-way ANOVA protocol, followed by Fisher's post hoc test in the acute assay. The three factors used in ANOVA were related to the polybrominated compounds (TBP and BDE-209), the dose (0.5, 5.0 and 50 ng g − 1 ) and the exposure time (24, 48, 72 and 96 h).
In the chronic experiment, the variables were analyzed using the one-way ANOVA protocol followed by Fisher's post hoc test. Additionally, the integrated biomarker response (IBR) index, described by Beliaeff and Burgeot (2002) and modi ed by Sanchez et al. (2013), was used to evaluate the sum of the effects of multibiomarkers (biochemical and histological) between the polybrominated compounds and doses.

Somatic biomarkers
The hepatosomatic index (HSI) was signi cantly in uenced by interaction between "polybrominated compounds" and "doses" (F 2, 61 = 4.15, p < 0.05). Only in 50 ng g − 1 a statistical difference was observed between the polybrominated compounds, with HSI being higher in sh treated with TBP (mean ± SD = 1.61 ± 0.14). Similar HSI were recorded to TBP, regardless of dose. Higher HSI values were found in specimens treated with 0.5 ng g − 1 of BDE-209 (mean ± SD = 1.65 ± 0.12) and lower in 50 ng g − 1 (mean ± SD = 1.38 ± 0.12), but no difference in GSI was found for both tested compounds.

Biochemical biomarkers
The AChE activity in the brain and muscle was higher in animals treated with TBP. Both biomarkers were in uenced by the interaction between "polybrominated compounds" and "doses" and by the association between "polybrominated" and "exposure time". At lower doses, AChE activity in the brain was greater in individuals treated with TBP. At 50 ng g − 1 , AChE activity in the brain increased signi cantly in sh exposed to BDE-209, in order equal to that observed in TBP (Fig. 1A). Only in groups exposed to BDE-209 an increase in AChE activity in the brain was observed according to the exposure time, mainly in 96h (Fig. 1B). The AChE activity in the muscle increased only in sh exposed to 5 ng g − 1 de TBP (Fig. 1C). After 72h of exposure, a signi cant increase in AChE activity in the muscle was found in animals exposed to TBP, regardless of dose. In the other exposure times, the AChE values in the muscle were similar between the polybrominated compounds (Fig. 1D).
The GST activity did not differ statistically in sh treated with TBP and BDE-209, being in uenced only by doses. While the lowest dose led to the highest GST activity, the intermediate and the highest doses led to smaller and similar activities (Fig. 1E). NPT and LPO levels were signi cantly higher in the animals treated with TBP, being in uenced by interaction between "polybrominated compounds" and "doses". The highest values of NPT occurred in individuals exposed to TBP, especially at the lowest dose (Fig. 1F). In 5.0 and 50 ng g − 1 TBP, NPT levels were reduced by half, equaling the levels observed in sh exposed to BDE-209. Higher levels of LPO were observed for both polybrominated compounds in individuals exposed to 0.5 and 5 ng g − 1 , decreasing at the highest dose (Fig. 1G). PCO levels were affected by the interaction of three factors: "polybrominated compounds", "dose" and "exposure time". In general, the animals presented a decrease of PCO levels with increasing dose. At 0.5 ng g − 1 TBP, the PCO values were lower at 24h of exposure and higher at 96h. Conversely, the highest PCO levels was recorded in 24h and the lowest in 96h in animals exposed to 5 ng g − 1 TBP. Individuals exposed to the highest dose of BDE-209 had higher levels of PCO in 24 h, decreasing after 96 h (Fig. 1H).

Chronic Experiment Somatic indexes
The somatic biomarkers were signi cantly in uenced by the polybrominated compounds and doses. The lowest HSI were recorded in BDE-209 treatments and in sh exposed to 5 and 50 ng g − 1 TBP, while GSI presented higher values only in individuals exposed to 0.5 ng g − 1 TBP. Condition factor was highest in control sh and lowest in BDE-209 treated, regardless of dose (Table 1).

Biochemical biomarkers
The AChE activity in the brain was signi cantly higher in animals exposed to BDE-209 and lower in those exposed to TBP, both at the dose of 50 ng g − 1 ( Fig. 2A). In the muscle, the AChE activity was higher only at the intermediate dose of  In the other treatments, AChE activity in the muscle was similar to the control group (Fig. 2B). The GST activity was higher in sh exposed to 50 ng g − 1 of BDE-209 and TBP than in the control group (Fig. 2C). The NPT concentration was higher only in individuals exposed to 50 ng g − 1 of BDE-209 in comparison to the control (Fig. 2D). Higher levels of LPO and PCO were observed in sh exposed to BDE-209, independent of the dose (Figs. 2E and F). Likewise, sh exposed to TBP had higher LPO at 50 ng g − 1 and PCO at 5 ng g − 1 than the control (Figs. 2E and F).

Histopathology
Eleven histopathological ndings were considered in the liver ( The Table 2 shows the incidence of changes and the lesion index, according to Bernet et al. (1999)   The bold values highlight the incidence of alteration, while the italic and underlined values highlight the tested concentration of histopathological ndings.
Comparatively, the hepatocytes of individuals exposed to TBP present distinct morphological aspects, with cholestasis widely found in the cytoplasm (Figs. 5d and E). Also, damages in the biliary canaliculi and the presence of vesicular deposits were observed (Fig. 5A). The cytoplasm was very affected showing a total disorganization (Fig. 5C) of organelles as rough endoplasmic reticulum fragmentation (Fig. 5B) and nuclear damages (Figs. 5D, E and F). It is quite di cult to identify other organelles, such as mitochondria. Cell death by necrosis was also found in the liver (Fig. 5E).
In general, the hepatocytes of individuals exposed to BDE-209 were also affected by the polybrominated compound. Cell death (Figs. 6A and F), a high presence of cholestasis (Fig. 6B) and focus of cytoplasm degeneration (Figs. 6A and D) were observed, as well as the presence of differentiated tissue like neoplastic focus (Fig. 6C), highlighting the effects on Disse's space and endothelial cells from sinusoids. Finally, damages in the blood vessels with necrosis increased the presence of hemorrhage (Fig. 6F).

Integrated analysis
A greater sum of responses to toxic effects was observed in sh exposed to BDE-209 (IBR = 4.78, Fig. 7B) compared to TBP (IBR = 4.08, Fig. 7A), mainly at 5 and 50 ng g − 1 due to the increase in NPT, LPO and PCO levels, and alteration of GST and AChE activities. In general, TBP presented a progressive increase in effects with the dose. LPO, NPT, PCO and histopathology levels, as well as the decrease of AChE activity in the brain were evident in sh exposed to 50 ng g − 1 of TBP (Fig. 7A).

Discussion
The current study shows that the polybrominated compounds TBP and BDE-209 present time-dependent toxic effects in adult male O. niloticus after acute and chronic exposures to environmentally relevant doses. In general, the biological disturbs were observed at different levels (biochemical, morphological and physiological) on target organs.
The acute and chronic experiments revealed that the tested doses of both brominated compounds led to disturbs on the size of the liver and gonads. The acute effects were related with higher doses of TBP (increase of HSI), whereas the lowest doses presented a more expressive effect on the liver (HSI) and gonads (GSI) after chronic exposure. Additionally, the lowest dose of TBP caused an increase of GSI. Conversely, individuals exposed to BDE-209 showed more sensitivity to lower doses in both acute and chronic exposures, with an increase of HSI in the acute exposure and a decrease of HSI in the chronic exposure. These ndings showed the importance to associate the acute and chronic studies to a better understanding of the toxic effects of chemicals, despite of these biological parameters being more useful for long-term of exposure. Recently, Folle et al. (2021) described that HSI was not in uenced by sex in O. niloticus exposed to TBP at similar doses, but GSI were more affected in female sh. In general, according to the somatic indexes TBP seems to be more toxic to liver in a time-dependent way since the same effects were observed only at high doses for acute exposure and at lower doses under chronic exposure. This nding is important because it represents a realistic condition of long-term exposure to low doses of polybrominated compounds that are naturally found in aquatic environments.
Both polybrominated compounds may be neurotoxic to O. niloticus after acute and chronic exposures. An increase of AChE activity was evident in the brain after acute exposure to the three doses of TBP and the highest dose of BDE-209. However, Xie et al. (2014) described a decrease of AChE activity after acute exposure (96 h) of juveniles Carassius auratus to much higher doses of BDE-209 (1-5 µg g − 1 ), using intraperitoneal injection. Effects on AChE activity can affect neuron cells communication in the central nervous system. Interestingly, the effects of TBP were observed even at low doses, while for BDE-209 they occurred at the highest dose and in a time-dependent way. According to Wang et al. (2018), BDE-209 interacts with AChE through hydrophobic interactions resulting in a change of the enzyme conformation, which may lead to its inhibition. Overcompensation through de novo synthesis of AChE, following a prior inhibition, may explain the increase of AChE activity observed after exposure to TBP and BDE-209. Comparatively, TBP seems to present a more neurotoxic effect in O. niloticus, including here a decrease of AChE activity observed in the chronic exposure, also described by Folle et al. (2021) for the same species. In Danio rerio, the cholinergic neurons are distributed in telencephalon, diencephalon, hypothalamus and mesencephalon (Clemente et al., 2004;Kaslin et al., 2004;Mueller et al., 2004), showing a high diversity of potential target regions of the brain sensitive to TBP and BDE-209 disturbs. Finally, the current study demonstrated that both polybrominated compounds can change the functional balance of neurotransmitter in brain, through alteration of AChE activity, that modulates electrical signal between neurons with consequences for central nervous system.
The effects of both brominated compounds in the liver were also in uenced by the duration of exposure, as the results showed a decrease in the activity of GST after acute exposure and an increase after chronic exposure, both at highest doses. The opposite effect was observed in C. auratus after acute exposure to higher doses of BDE-209 (Xie et al., 2014) and by Folle et al. (2021) after chronic exposure. These ndings strongly suggest that the effects of the brominated compounds TBP and BDE-209 are highly in uenced by the exposure period under environmental doses, whereas the effects at short-term exposure are dose-dependent. GST is a biotransformation enzyme extensively used as biomarker to evaluate the effects of pollutants in sh. The decrease of the activity in acute exposure suggests a disturb in the hepatocyte detoxi cation mechanism. The increase of GST activity after chronic exposure can also be a response to a pro-oxidant status that can explain the high level of damages in lipids (LPO) and proteins (PCO), and so con rming an oxidative stress in the cells. Xie et al. (2014) described the inhibition of catalase activity after exposure to BDE-209 supporting the oxidative stress as discussed above.
Despite of that, the GST also play an important role in protecting tissues against xenobiotics and electrophilic substances. According to the current data, the acute effects of both tested compounds mean a hepatocyte sensitization, while the chronic exposure shows a late activation of detoxi cation mechanism that was not enough to protect the cells, as observed. Ghosh et al. (2013) also described an increase of GST activity after chronic exposure to PBDE mixtures, although it is also involved with debromination (Roberts et al., 2011).
Morphological damages, such as histopathological and ultrastructural ndings in tissues and cells represent effects of long-term exposure to environmental relevant doses of chemicals due to failure of protective mechanisms (Gemuse et al., 2021). The occurrence of histopathological ndings showed distinct effects depending on the compound, period of exposure and doses.
The occurrence of necrosis, a degenerative damage, was found in all studied groups including the control. This kind of effect reveals the failure of defense mechanisms in cells or tissues and represent a permanent effect that depending on the extension can decrease the liver function (Folle et al., 2021). Necrosis is more related with chronic exposure and in the current study individuals exposed to BDE-209 showed higher values comparatively with those exposed to TBP. There was a dose-depend effect for TPB, but not to BDE-209. For acute exposure, only the individuals exposed to TPB presented dose and time-dependent effects that were not observed to individuals from the groups exposed to BDE-209. On this way, BDE-209 was more toxic to O. niloticus than TBP in terms of necrosis incidence.
The presence of cholestasis means a de ciency in hepatocyte bile metabolism or secretion. In general, the accumulation of some toxic bile salts in the cell cytoplasm can cause cell death by necrosis and apoptosis (Boelsterli, 2007). This nding can also be explained by obstruction of bile ow in biliary ducts, which was observed in the ultrastructure analysis. Individuals exposed to BDE-209 showed damages in the biliary duct organization and sinusoids structure, reinforcing the occurrence of cholestasis and, consequently, the permanent damages observed in the study, like necrosis. Additionally, according to Ho et al. (2012) the metabolism of TBP can produce a more toxic compound, the tribromoanisole, through the action of methyltransferases, which can also explain the permanent liver damages found in the current study.
In ammation is a pharmacological event that can be identi ed by the occurrence of leucocytes in ltration, perivascular or peritubular granulomatosis and melanomacrophage centers. These events were present in individuals exposed to both contaminants, but were more evident in groups after chronic exposure. In general, dose-dependent in ammatory responses occurred only after exposure to BDE-209, and it was corroborated by the results of Santana et al. (2018).
The in ammation in vessels can favor other events, like vascular congestion and hemorrhage, which were more common in individuals chronically exposed to BDE-209. In particular, vascular damages can compromise the liver function, since the role of liver depends on the constant and intense blood ow. The circulatory disturbs were also described by Bbarja-Fernández et al. (2013) in hepatic tissue of sh after chronic exposure to BDE-47.
The integrated biomarker analysis is an important tool to be considered and to understand the effects and consequence of exposure to pollutants. According to Xie et al. (2014), IBR might be a useful tool for quanti cation of integrated biological effects induced by coexisting contaminants toward sh. According to the acute experiment, TBP presented greater toxicity for O. niloticus than BDE-209, evidenced by the neurotoxic disturbs and high levels of lipid peroxidation. In this experiment, the intermediate dose accumulated more negative responses in the biomarkers. Additionally, the exposure time had a low in uence on the response pattern of the biomarkers. In the chronic experiment, this pattern changed, as BDE-209 presented a more expressive toxicity to O. niloticus than TBP.
The highest dose caused more negative effects on both polybrominated compounds, although for BDE-209 the doses of 0.5 and 5.0 ng g − 1 were also quite toxic for the animals.
In general, the effects observed after acute exposure were less evident than those observed for the chronic exposure. Also, effects usually occurred for the highest dose of TBP and the intermediate and highest doses of BDE-209. The ultrastructural analysis revealed that the hepatocytes were very affected by the chronic exposure to both contaminants and from acute and chronic exposure. This approach allows to observe effects at subcellular level that potentially explain the tissue damages observed by histopathological analysis. Both contaminants caused cholestasis represented by the accumulation of bile salt deposits as dark bodies inside the cells, which can lead to cell death, as observed by the high incidence of necrosis and cytoplasmic damages. The lesions in the endothelial cells indicate the easy access ingested pollutants have to the hepatic parenchyma. Damages in the Disse's space con rm the toxic effects of pollutants and explain the occurrence of perivascular granulomatosis events that can evolve to hemorrhagic or vascular congestion.

Conclusion
Exposure to environmentally relevant doses of TBP and BDE-209 revealed important ndings in O. niloticus. First, the effects seem to be more dependent on the dose for acute exposure, and less dependent on it for chronic exposure. Second, TBP seems to be more toxic than BDE-209, since effects were not restricted to the highest dose of the compound. Third, there is indication of neurotoxicity, and strong evidences of redox unbalance and toxicity to the liver.
Finally, the multi-biomarker approach performed here proved to be e cient as histopathology, ultrastructure and biochemical data con rm the risk of exposure of sh to both TBP and BDE-209.  Brazilian agencies CAPES (Coordination for the Improvement of Higher Education Personnel -Finance Code 001). This agency is responsible to institutional fellowship to the students develop the project (Dandie A Bozza).

Declarations
-Competing Interests: The authors declare that they have no competing interests -Availability of data and materials: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Biochemical biomarkers in the acute experiment. O. niloticus were exposed to 0.5, 5 and 50 ng g-1 of TBP and BDE-209. A-B: Acetylcholinesterase activity in the brain. C-D: Acetylcholinesterase activity in the muscle. E: Glutathione S-transferase activity. F: Non-protein thiols concentration. G: Lipid peroxidation levels.
H: Protein carbonylation levels. The ANOVA statistic data is summarized in the title of the gures. The bars indicate the mean ± con dence interval. The different letters on the bars indicate the signi cance of the Fisher test.

Figure 2
Biochemical biomarkers in the chronic experiment. O. niloticus were exposed to 0.5, 5 and 50 ng g-1 of TBP and BDE-209. A-B: Acetylcholinesterase activity in the brain and muscle. C: Glutathione S-transferase activity. D: Non-protein thiols concentration. E: Lipid peroxidation levels. F: Protein carbonylation levels. The ANOVA statistic is summarized in the title of the gures. The bars indicate mean ± con dence interval. The different letters on the bars indicate the signi cance of the Fisher test. Histopathology Figure 6 Ultrastructure of O. niloticus hepatocytes from sh exposed to BDE-209. A. The large arrow indicates a necrotic cell around the biliary canaliculum (small arrow). Observe the presence of cytoplasm degeneration (dl). B. The arrows indicate the presence of coletasis. C. Observe the presence of atypical cells (*). D.
Observe the lesions in the sinusoide (Sn) with the presence of damages in the endothelial cells (grey arrows) and an enlargement (black arrow) of Disse's space (white arrows). E. The black arrow indicates a biliary canaliculum degeneration with a pre-necrotic cell around it (white arrow). F. The white arrow indicates a necrotic cell around a degenerated vessel (grey arrow) leading to hemorragic process (black arrow) and red blood cells invasion (h). Bars = 1 µm.