Effects of Two Sublethal Concentrations of Mercury Chloride on the Morphology and Metallothionein Activity in the Liver of Zebrafish (Danio rerio)

Mercury (Hg) is a highly hazardous pollutant widely used in industrial, pharmaceutical and agricultural fields. Mercury is found in the environment in several forms, elemental, inorganic (iHg) and organic, all of which are toxic. Considering that the liver is the organ primarily involved in the regulation of metabolic pathways, homeostasis and detoxification we investigated the morphological and ultrastructural effects in Danio rerio liver after 96 h exposure to two low HgCl2 concentrations (7.7 and 38.5 μg/L). We showed that a short-term exposure to very low concentrations of iHg severely affects liver morphology and ultrastructure. The main effects recorded in this work were: cytoplasm vacuolization, decrease in both lipid droplets and glycogen granules, increase in number of mitochondria, increase of rough endoplasmic reticulum and pyknotic nuclei. Pathological alterations observed were dose dependent. Trough immunohistochemistry, in situ hybridization and real-time PCR analysis, the induction of metallothionein (MT) under stressor conditions was also evaluated. Some of observed alterations could be considered as a general response of tissue to heavy metals, whereas others (such as increased number of mitochondria and increase of RER) may be considered as an adaptive response to mercury.


Introduction
Mercury (Hg) is a naturally occurring metal derived from several processes, such as weathering of the earth's crust, soil erosion, geologic activities, and its contents in the environment can vary between different regions [1]. Human sources of mercury pollution include activities such as agriculture, municipal wastewater discharges, mining, incineration, and discharges of industrial wastewater [2].
These primary sources both natural and anthropogenic, have been directly mobilized by humans for thousands of years into aquatic and terrestrial ecosystems [3] resulting in an increased release into the environment that often exceed permissible levels.
Mercury has drawn the attention of scientific community since the 1950s when the first evidences of dangerous environmental effects have been reported [4], thus prompting policies and regulations to Hence in order to have a more comprehensive overview of iHg-induced effects, we evaluated histology, ultrastructure and MTs expression in Danio rerio liver after acute exposure to two low HgCl 2 concentrations with the aim of further elucidating the role of morphofunctional studies in monitoring aquatic environment.

Control
The morphology and ultrastructure of Danio rerio liver have been previously described [46,47] and only a brief general description relevant to the present work will be given. Danio rerio liver is a dense organ divided into three lobes, located in the cranial part of abdominal cavity surrounding the digestive tract. In this species hepatic lobules, portal areas and hepatic arteries that are characteristics of liver, cannot be observed. The blood enters in the liver trough the portal vein that branches directly into the sinusoids. The sinusoids converged to the central vessel and the blood is carried out of liver.
Under light microscope, liver shows a homogeneous parenchyma composed of hepatocytes ( Figure 1A) surrounding a network of sinusoids. Hepatocytes that surround the sinusoids are disposed in two rows that give rise to cords ( Figure 1B); the sinusoidal lumen is lined by a continuous endothelium where no basal lamina occurs ( Figure 1B). The parenchyma is well-drained by a reticulum of circular structures called bile duct that are enclosed by cuboidal epithelium ( Figure 1C). The space of Disse (also called perisinusoidal space) extends between the sinusoidal endothelium and the hepatocytes ( Figure 1B). Hepatocytes show uniform size and contain a big nucleus located in the center of the cell; a dense nucleolus is usually evident ( Figure 1B,D). Within the cytoplasm lipid droplets, often surrounded by numerous glycogen granules, could be observed ( Figure 1D). Hence in order to have a more comprehensive overview of iHg-induced effects, we evaluated histology, ultrastructure and MTs expression in Danio rerio liver after acute exposure to two low HgCl2 concentrations with the aim of further elucidating the role of morphofunctional studies in monitoring aquatic environment.

Control
The morphology and ultrastructure of Danio rerio liver have been previously described [46,47] and only a brief general description relevant to the present work will be given. Danio rerio liver is a dense organ divided into three lobes, located in the cranial part of abdominal cavity surrounding the digestive tract. In this species hepatic lobules, portal areas and hepatic arteries that are characteristics of liver, cannot be observed. The blood enters in the liver trough the portal vein that branches directly into the sinusoids. The sinusoids converged to the central vessel and the blood is carried out of liver.
Under light microscope, liver shows a homogeneous parenchyma composed of hepatocytes ( Figure 1A) surrounding a network of sinusoids. Hepatocytes that surround the sinusoids are disposed in two rows that give rise to cords ( Figure 1B); the sinusoidal lumen is lined by a continuous endothelium where no basal lamina occurs ( Figure 1B). The parenchyma is well-drained by a reticulum of circular structures called bile duct that are enclosed by cuboidal epithelium ( Figure  1C). The space of Disse (also called perisinusoidal space) extends between the sinusoidal endothelium and the hepatocytes ( Figure 1B). Hepatocytes show uniform size and contain a big nucleus located in the center of the cell; a dense nucleolus is usually evident ( Figure 1B,D). Within the cytoplasm lipid droplets, often surrounded by numerous glycogen granules, could be observed ( Figure 1D). Observed under TEM, the space of Disse appears as a region of low electron-density ( Figure  2A).
The cytoplasm of the hepatocytes shows an high grade of electron-density and it is characterized by a well-developed rough endoplasmic reticulum (RER) that often runs parallel to the cellular membrane and nuclear membrane (Figure 2A). Numerous mitochondria of different shape (spherical and oval forms) are scattered through the cytoplasm (Figure 2A,B); they show few cristae and some electron-dense glycogen granules. Golgi complex is poorly developed and smooth endoplasmic reticulum (SER) is not appreciable. A large amount of low density glycogen granules are mainly gathered at the periphery of the hepatic cytoplasm ( Figure 2B); several small lipid Observed under TEM, the space of Disse appears as a region of low electron-density ( Figure 2A). The cytoplasm of the hepatocytes shows an high grade of electron-density and it is characterized by a well-developed rough endoplasmic reticulum (RER) that often runs parallel to the cellular membrane and nuclear membrane (Figure 2A). Numerous mitochondria of different shape (spherical and oval forms) are scattered through the cytoplasm (Figure 2A,B); they show few cristae and some electron-dense glycogen granules. Golgi complex is poorly developed and smooth endoplasmic reticulum (SER) is not appreciable. A large amount of low density glycogen granules are mainly gathered at the periphery of the hepatic cytoplasm ( Figure 2B); several small lipid droplets of different size can also be observed (Figure 2A,B). Nucleus is spherical or ovoid and is usually located in the center of the hepatocyte; the granular euchromatin occupies the most part of the nucleus whereas the heterochromatin is distributed at the periphery; an highly stained nucleolus is also observed, often in the central region of the nucleus (Figure 2A,B). droplets of different size can also be observed (Figure 2A,B). Nucleus is spherical or ovoid and is usually located in the center of the hepatocyte; the granular euchromatin occupies the most part of the nucleus whereas the heterochromatin is distributed at the periphery; an highly stained nucleolus is also observed, often in the central region of the nucleus (Figure 2A,B).

Exposed Fishes
After 96 h of exposure to 7.7 μg/L of HgCl2 (for simplicity, we will refer to 7.7 μg/L as the "low" concentration), the histological observation reveals severe alteration in liver morphology of treated animals compared with control ( Figure 3A). The liver parenchyma undergoes degenerative changes and it is possible to observe large areas of lysis ( Figure 3B) whereas sinusoids and venules are filled with red blood cells ( Figure 3C).
The major changes observed in the hepatic parenchyma of exposed fishes include an intense cytoplasm vacuolization and a decrease in the content of both lipid droplets and glycogen granules ( Figure 3C). Ultrastructural observations reveal the modification of the Disse's space architecture with alterations to the endothelial cell and congestion of blood vessels ( Figure 4A). TEM observations confirm the conspicuous vacuolization of the hepatocytes cytoplasm that shows a wide

Exposed Fishes
After 96 h of exposure to 7.7 µg/L of HgCl 2 (for simplicity, we will refer to 7.7 µg/L as the "low" concentration), the histological observation reveals severe alteration in liver morphology of treated animals compared with control ( Figure 3A). The liver parenchyma undergoes degenerative changes and it is possible to observe large areas of lysis ( Figure 3B) whereas sinusoids and venules are filled with red blood cells ( Figure 3C).
The major changes observed in the hepatic parenchyma of exposed fishes include an intense cytoplasm vacuolization and a decrease in the content of both lipid droplets and glycogen granules ( Figure 3C). droplets of different size can also be observed (Figure 2A,B). Nucleus is spherical or ovoid and is usually located in the center of the hepatocyte; the granular euchromatin occupies the most part of the nucleus whereas the heterochromatin is distributed at the periphery; an highly stained nucleolus is also observed, often in the central region of the nucleus (Figure 2A,B).

Exposed Fishes
After 96 h of exposure to 7.7 μg/L of HgCl2 (for simplicity, we will refer to 7.7 μg/L as the "low" concentration), the histological observation reveals severe alteration in liver morphology of treated animals compared with control ( Figure 3A). The liver parenchyma undergoes degenerative changes and it is possible to observe large areas of lysis ( Figure 3B) whereas sinusoids and venules are filled with red blood cells ( Figure 3C).
The major changes observed in the hepatic parenchyma of exposed fishes include an intense cytoplasm vacuolization and a decrease in the content of both lipid droplets and glycogen granules ( Figure 3C). Ultrastructural observations reveal the modification of the Disse's space architecture with alterations to the endothelial cell and congestion of blood vessels ( Figure 4A). TEM observations confirm the conspicuous vacuolization of the hepatocytes cytoplasm that shows a wide Ultrastructural observations reveal the modification of the Disse's space architecture with alterations to the endothelial cell and congestion of blood vessels ( Figure 4A). TEM observations confirm the conspicuous vacuolization of the hepatocytes cytoplasm that shows a wide electron-clear area ( Figure 4B). It is also evident the dilation and fragmentation of RER cisternae and the decrease of glycogen reserves ( Figure 4B,C). Nuclei often lose their rounded shape and show a change in both distribution and amount of heterochromatin ( Figure 4B). Cytoplasmic organelles are no longer distinguishable ( Figure 4B,C). electron-clear area ( Figure 4B). It is also evident the dilation and fragmentation of RER cisternae and the decrease of glycogen reserves ( Figure 4B,C). Nuclei often lose their rounded shape and show a change in both distribution and amount of heterochromatin ( Figure 4B). Cytoplasmic organelles are no longer distinguishable ( Figure 4B,C). After 96 h of exposure to 38.5 μg/L of HgCl2 (for simplicity, we will refer to 38.5 μg/L as the "high" concentration), the liver morphology appears markedly altered. ( Figure 5A) In the liver parenchyma, conspicuous areas of lysis are evident ( Figure 5B) along with vessels congestion ( Figure  5C).
In the hepatic parenchyma of exposed fishes, cytoplasm vacuolization is not detected and the most evident feature is the presence of a large and extended RER at the periphery of the nucleus ( Figure 5D). A decrease in the content of both lipid droplets and glycogen granules can also be detected ( Figure 5C,D). Within the parenchyma several degenerate hepatocytes are observed ( Figure  5D). Under TEM, it is possible to detect the morphological alterations of Disse's space, and presence of numerous red blood cells filling sinusoids and venules ( Figure 6A). Ultrastructural observations provided further evidence of the great development of RER which occupies large parts of the electron-clear cytoplasm of hepatocytes ( Figure 6B-F). A conspicuous Golgi apparatus with curved cisternae can also be seen ( Figure 6B). The numerous sparse mitochondria show a slightly increased electron-density ( Figure 6B,C). In the central portion of the nucleus a conspicuous nucleolus can After 96 h of exposure to 38.5 µg/L of HgCl 2 (for simplicity, we will refer to 38.5 µg/L as the "high" concentration), the liver morphology appears markedly altered. ( Figure 5A) In the liver parenchyma, conspicuous areas of lysis are evident ( Figure 5B) along with vessels congestion ( Figure 5C). electron-clear area ( Figure 4B). It is also evident the dilation and fragmentation of RER cisternae and the decrease of glycogen reserves ( Figure 4B,C). Nuclei often lose their rounded shape and show a change in both distribution and amount of heterochromatin ( Figure 4B). Cytoplasmic organelles are no longer distinguishable ( Figure 4B,C). After 96 h of exposure to 38.5 μg/L of HgCl2 (for simplicity, we will refer to 38.5 μg/L as the "high" concentration), the liver morphology appears markedly altered. ( Figure 5A) In the liver parenchyma, conspicuous areas of lysis are evident ( Figure 5B) along with vessels congestion ( Figure  5C).
In the hepatic parenchyma of exposed fishes, cytoplasm vacuolization is not detected and the most evident feature is the presence of a large and extended RER at the periphery of the nucleus ( Figure 5D). A decrease in the content of both lipid droplets and glycogen granules can also be detected ( Figure 5C,D). Within the parenchyma several degenerate hepatocytes are observed ( Figure  5D). Under TEM, it is possible to detect the morphological alterations of Disse's space, and presence of numerous red blood cells filling sinusoids and venules ( Figure 6A). Ultrastructural observations provided further evidence of the great development of RER which occupies large parts of the electron-clear cytoplasm of hepatocytes ( Figure 6B-F). A conspicuous Golgi apparatus with curved cisternae can also be seen ( Figure 6B). The numerous sparse mitochondria show a slightly increased electron-density ( Figure 6B,C). In the central portion of the nucleus a conspicuous nucleolus can In the hepatic parenchyma of exposed fishes, cytoplasm vacuolization is not detected and the most evident feature is the presence of a large and extended RER at the periphery of the nucleus ( Figure 5D). A decrease in the content of both lipid droplets and glycogen granules can also be detected ( Figure 5C,D). Within the parenchyma several degenerate hepatocytes are observed ( Figure 5D). Under TEM, it is possible to detect the morphological alterations of Disse's space, and presence of numerous red blood cells filling sinusoids and venules ( Figure 6A). Ultrastructural observations provided further evidence of the great development of RER which occupies large parts of the electron-clear cytoplasm of hepatocytes ( Figure 6B-F). A conspicuous Golgi apparatus with curved cisternae can also be seen ( Figure 6B). The numerous sparse mitochondria show a slightly increased electron-density ( Figure 6B,C). In the central portion of the nucleus a conspicuous nucleolus can often be seen whereas heterochromatin concentrates at the periphery ( Figure 6B-F). Scattered through the liver parenchyma several degenerating hepatocytes can be detected ( Figure 6C). At several point a swollen RER cisternae, and pyknotic nuclei are evident ( Figure 6B,C). In the cytoplasm, atypical cytoplasmic electron-dense granules are often detected ( Figure 6C,E). often be seen whereas heterochromatin concentrates at the periphery ( Figure 6B-F). Scattered through the liver parenchyma several degenerating hepatocytes can be detected ( Figure 6C). At several point a swollen RER cisternae, and pyknotic nuclei are evident ( Figure 6B,C). In the cytoplasm, atypical cytoplasmic electron-dense granules are often detected ( Figure 6C,E).

In Situ Hybridization
In the control group the hybridization signal is too weak to allow confident identification of the cells expressing mt mRNA ( Figure 7A).
Compared to the control group, a positive hybridization signal is observed in liver of specimens exposed to the low concentration. In particular, a marked increase in signal development in the whole parenchyma is detected with a peak in the cytoplasmic areas of the hepatocytes ( Figure 7B). In specimens from the high concentration group, the transcript observed in the liver parenchyma shows a slightly higher intensity compared to that detected in the low concentration group: the signal is observed in all liver tissue and is more intense at the periphery of parenchyma ( Figure 7C).

In Situ Hybridization
In the control group the hybridization signal is too weak to allow confident identification of the cells expressing mt mRNA ( Figure 7A).
Compared to the control group, a positive hybridization signal is observed in liver of specimens exposed to the low concentration. In particular, a marked increase in signal development in the whole parenchyma is detected with a peak in the cytoplasmic areas of the hepatocytes ( Figure 7B). In specimens from the high concentration group, the transcript observed in the liver parenchyma shows a slightly higher intensity compared to that detected in the low concentration group: the signal is observed in all liver tissue and is more intense at the periphery of parenchyma ( Figure 7C).

Immunolocalization of MTs
In the liver of animals from control group very few labeled cells show a weak immunoreactivity for MTs ( Figure 8A). After 96 h of exposure to the low concentration, the immunohistochemical analysis reveal a significant increase in MTs labeling; a pronounced MTs immunoreactivity can be observed in the whole liver parenchyma ( Figure 8B). After 96 h of exposure to the high concentration, MTs labeling increases; immunoreactive cells are distributed throughout the liver parenchyma ( Figure 8C).

Immunolocalization of MTs
In the liver of animals from control group very few labeled cells show a weak immunoreactivity for MTs ( Figure 8A). After 96 h of exposure to the low concentration, the immunohistochemical analysis reveal a significant increase in MTs labeling; a pronounced MTs immunoreactivity can be observed in the whole liver parenchyma ( Figure 8B). After 96 h of exposure to the high concentration, MTs labeling increases; immunoreactive cells are distributed throughout the liver parenchyma ( Figure 8C).

In Situ Hybridization
In the control group the hybridization signal is too weak to allow confident identification of the cells expressing mt mRNA ( Figure 7A).
Compared to the control group, a positive hybridization signal is observed in liver of specimens exposed to the low concentration. In particular, a marked increase in signal development in the whole parenchyma is detected with a peak in the cytoplasmic areas of the hepatocytes ( Figure 7B). In specimens from the high concentration group, the transcript observed in the liver parenchyma shows a slightly higher intensity compared to that detected in the low concentration group: the signal is observed in all liver tissue and is more intense at the periphery of parenchyma ( Figure 7C).

Immunolocalization of MTs
In the liver of animals from control group very few labeled cells show a weak immunoreactivity for MTs ( Figure 8A). After 96 h of exposure to the low concentration, the immunohistochemical analysis reveal a significant increase in MTs labeling; a pronounced MTs immunoreactivity can be observed in the whole liver parenchyma ( Figure 8B). After 96 h of exposure to the high concentration, MTs labeling increases; immunoreactive cells are distributed throughout the liver parenchyma ( Figure 8C).

Gene Expression
The genetic analysis evidenced a tissue-specific expression rate. In addition to the basal genetic expression of mt in liver, the gene shows a significant modification of its expression in both treatments after 96 h of exposure to HgCl 2 (Figure 9). Mt gene is significantly increased (p < 0.001) at 7.7 µg/L exposure level compared to control group; the gene shows the highest responses at 38.5 µg/L of HgCl 2 when compared to basal condition (p < 0.001). The genetic analysis evidenced a tissue-specific expression rate. In addition to the basal genetic expression of mt in liver, the gene shows a significant modification of its expression in both treatments after 96 h of exposure to HgCl2 (Figure 9). Mt gene is significantly increased (p < 0.001) at 7.7 μg/L exposure level compared to control group; the gene shows the highest responses at 38.5 μg/L of HgCl2 when compared to basal condition (p < 0.001). Figure 9. Relative variations in mt Gene Expression As Compared to control in Danio rerio liver after 96 h of exposure to HgCl2 (7.7-38.5 μg/L). Taqman real time relative quantitative PCR. The bars show mean ± S.D, n = 5. Asterisks indicate the treated groups that differ from the control, *** p < 0.001; hashtags indicate difference between treated groups ### p < 0.001 (One way ANOVA followed by Bonferroni's post hoc test).

Discussion
Our observation on healthy Danio rerio, confirm the previously described liver organization [46,47]; the homogeneous parenchyma is composed of hepatocytes, arranged in cords, surrounding a network of sinusoids that converged to the central vessel. Hepatic lobules, portal areas, and hepatic arteries were not seen.
In the present study we investigated the effects of HgCl2 on zebrafish liver parenchyma through a morphofunctional and ultrastructural study; we showed that liver injury was induced when fishes were exposed to mercury chloride for 96 h at doses of 7.7 and 38.5 μg/L; the doses of mercury administered in our experiments were very low, but high enough to produce severe liver injury and also inducing an enhanced expression of hepatic MTs. These effects was dose dependent.
Examination of liver after exposure to both tested concentrations showed remarkable effects, resulting in modifications in both histology and ultrastructure.
The dilation of Disse's space and the occurrence of wide lysed areas were the most evident result of Hg application. Morphological disorders and necrosis in the Disses's space were found in Hoplias malabaricus after chronic exposure to dietary methylmercury [30]. These authors suggested that endothelial cells alterations may affect the exchange between hepatocytes and sinusoids (e.g., nutrients, chemical signals and physiological components) thus resulting in the disturbance of cell physiology and organ function stability.
The appearances of lysed areas were found in zebrafish liver after exposure to copper sulfate [48]. Considering the copper's ability to induce membrane disruption, these authors suggested that the lysis distribution may reflect a heterogeneous distribution of Cu in the parenchyma. Interestingly, we observed the appearance of lysed area after exposure to iHg; whereas similar adverse effects have not been reported in liver after exposure to MeHg; these results further support

Discussion
Our observation on healthy Danio rerio, confirm the previously described liver organization [46,47]; the homogeneous parenchyma is composed of hepatocytes, arranged in cords, surrounding a network of sinusoids that converged to the central vessel. Hepatic lobules, portal areas, and hepatic arteries were not seen.
In the present study we investigated the effects of HgCl 2 on zebrafish liver parenchyma through a morphofunctional and ultrastructural study; we showed that liver injury was induced when fishes were exposed to mercury chloride for 96 h at doses of 7.7 and 38.5 µg/L; the doses of mercury administered in our experiments were very low, but high enough to produce severe liver injury and also inducing an enhanced expression of hepatic MTs. These effects was dose dependent.
Examination of liver after exposure to both tested concentrations showed remarkable effects, resulting in modifications in both histology and ultrastructure.
The dilation of Disse's space and the occurrence of wide lysed areas were the most evident result of Hg application. Morphological disorders and necrosis in the Disses's space were found in Hoplias malabaricus after chronic exposure to dietary methylmercury [30]. These authors suggested that endothelial cells alterations may affect the exchange between hepatocytes and sinusoids (e.g., nutrients, chemical signals and physiological components) thus resulting in the disturbance of cell physiology and organ function stability.
The appearances of lysed areas were found in zebrafish liver after exposure to copper sulfate [48]. Considering the copper's ability to induce membrane disruption, these authors suggested that the lysis distribution may reflect a heterogeneous distribution of Cu in the parenchyma. Interestingly, we observed the appearance of lysed area after exposure to iHg; whereas similar adverse effects have not been reported in liver after exposure to MeHg; these results further support the hypothesis that iHg, affect membrane whereas methylmercury does not appear to act directly at plasma membrane [22,49].
In the present study, a drastic decrease in glycogen amount and lipid granules was observed in the hepatocyte of both low and high exposed groups.
Glycogen depletion has been reported in liver fishes, after short and long term exposure to other heavy metals, such as copper, cadmium and lead [48,50,51]. Moreover, similar glycogen perturbations have also been recorded in both liver and muscle of Oreochromis mossambicus as result of mercury chloride toxicity; Roy George and colleagues [27] suggested that the great loss of glycogen in the liver indicates that it is the most affected organ during stress. Since the liver is the main site for detoxification, glycogen depletion could be correlated with the increased glycolytic activity required by enhanced metabolism under stress conditions.
In the hepatocytes the presence of glycogen granules and lipid droplet is related to the normal metabolic function of the cells. In addition, the depletion of lipids described in our work is often observed in fishes after exposure to both organic and inorganic mercury [31,52]. The depletion of both reserve substances may suggest an increase of basal metabolic pathways of hepatocytes, due to the histological modification after experimental conditions. After exposure to low HgCl 2 concentration, other common ultrastructural alterations observed in zebrafish hepatocytes were: organelle disorganization within cytoplasm (empty appearance), cytoplasmic vacuolization, RER dilation and fragmentation. Similar morphological alterations were also found after exposure to other heavy metal suggesting an unspecific adaptive response of the liver to stress [48]. Interestingly, some ultrastructural alterations observed after the exposure to high Hg concentration such as increased number of mitochondria and increase of RER, seems to be adaptive. The great development of RER indicate an enhanced hepatic protein synthesis [48].
The electron dense granule in the cytoplasm along with the increase of heterochromatin in hepatocytes nuclei have been reported after exposure to mercury in Hoplias malabaricus. The authors suggested that this morphological alteration of nuclei is evidence that this organelle can accumulate metals more intensely than other cellular compartments [30].
Numerous studies have demonstrated how the metallothioneins are important in the regulation of biological functions such as homeostasis of zinc and copper [53,54], metal detoxification [55,56] and oxygen radical scavenging [57,58]. The MTs have been elected in both vertebrates and invertebrates as a biomarker of environmental metal pollution [59][60][61][62]. Based on the strong correlation between MTs expression and environmental concentrations of metals, the amount of MTs in the cells is considered essential for monitoring the exposure effects of heavy metals and for predicting their toxic effects on the cells [63,64].
Our functional and molecular results have showed that the immunolocalization of MT protein and the expression of mt mRNA were induced after exposure to HgCl 2 in Danio rerio liver in a dose dependent way. The results of this work showed that the induction of MTs after exposure to the contaminant was significant for both experimental concentrations compared to the control group. Furthermore, the functional investigation and the in situ hybridization analysis revealed that the MT protein and the mt mRNA were located in the cytoplasm of hepatocytes. The MTs are usually localized in the cytoplasm of the cells, but can be transferred into the nucleus according to the cell cycle changes [65]. Analysis of the localization of both mt mRNA and MT protein is important to clarify the physiological function of MTs [66]. According to Woo and Lazo [67] the cytoplasm localization of MTs has a suppressive effect on the cytotoxicity induced by heavy metals and decreases the level of intracellular reactive oxygen.
Ivanina and his collaborators [68] suggested that the metal toxicity in the cells is greatly reduced once the toxicant bound the MTs. This kind of connection between the MTs and the metals prevents the heavy metal interaction with other cellular components such as enzymes, proteins, DNA and lipids.
Furthermore, high levels of MTs give to the cell sufficient protection from the toxic effects of heavy metals such as oxidation, denaturation and misfolding of protein.
In our study, although the increased expression of MTs in the liver, histological and ultrastructural damage was severe thus suggesting that the high toxicity of mercury does not allow MTs to perform successfully the metal detoxification in the cells.
In this work functional changes have been correlated to morphological and ultrastructural perturbations induced in liver of Danio rerio after exposure to mercury chloride.
Our results provide valuable information for a more comprehensive understanding of the effects of inorganic mercury on fish liver supporting the strong relationship between morphological and functional biomarkers. Further studies are needed to better understand the dangerous effects of heavy metals on hepatic tissue in order to have a more comprehensive overview of the impact on aquatic biota.

Test Organism and Experimental Design
Danio rerio (84 healthy adults of both sexes of length 0.35˘0.5 cm and weight 0.43˘0.06 g) were purchased from a local shop of ornamental fishes. Prior to experiments, fishes were acclimatized to laboratory conditions for 2 weeks in two 80 L aquaria filled with continuously aerated dechlorinated water. The temperature was kept at 25˘0.5˝C under a 12/12-h light-dark photoperiod. The fishes were fed every 2 days during the acclimation period with commercial fish food. Zebrafish were exposed in a static system for 96 h to two mercury chloride (HgCl 2 ; Sigma-Aldrich Chemical Co., St. Louis, MO, USA) concentrations of 7.7 and 38.5 µg/L; these concentrations correspond to 10% and 50% of the median lethal concentration at 96 h (LC50 96 ) [69]. For each experimental unit, 14 animals of comparable body dimension were randomly chosen and assigned to 30 L constantly aerated glass tanks. Animals from the control group were kept in aged tap water. During experimental period animals were not fed and temperature, photoperiod, and aeration were kept under the same conditions used during the acclimation. Each treatment was conducted in duplicate.
The liver was removed after 96 h of exposure. Fishes from both control and exposed groups were anesthetized with tricaine methane sulfonate MS 222 (Sigma-Aldrich Chemical Co., St. Louis, MO, USA). Animal care, killing and experiments were supervised according to the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes (Council of Europe No. 123, Strasbourg, 1985).

Light Microscopy and Transmission Electron Microscopy
The excised livers, cut into small pieces, were immediately fixed by direct immersion in 4% glutaraldehyde (Electron Microscopy Sciences, Hatfield, PA, USA) in phosphate-buffered saline (PBS 0.1 M, pH 7.2) for 48 h at 4˝C. After post-fixation for 2 h with osmium tetroxide (1% in the same buffer) and dehydration in graded ethanol, specimens for light microscopy (LM) and transmission electron microscopy (TEM) were embedded in epoxy resin (Araldite 502/Embed 812, Electron Microscopy Sciences). Semi-thin sections (1 µm) were stained with toluidine blue; these sections were observed and photographed by a LM LEITZ Dialux EB 20 (Leica Microsystems, Wetzlar, Germany). Ultrathin sections (800 Å) were stained with uranyl acetate replacement and contrasted using lead citrate (Electron Microscopy Sciences), then observed by a JEOL 1011 electron microscope (JEOL, Inc., Peabody, MA, USA).

Immunohistochemistry
For the immunohistochemical study, samples were fixed in Bouin solution for 48 h and then dehydrated in an increasing series of ethanol. The samples were xylene-cleared and embedded in paraffin wax (mean fusion point of 56˝C). Tissue sections (8 µm) were cut and mounted on slides. On deparaffinized and hydrated sections it was applied the indirect immunofluorescence technique [70]. To block non-specific sites, the slides were washed three times in PBS (pH 7.4) and then the sections were incubated for 10 min in a moist chamber with 20% normal goat serum. Unwashed slides were incubated overnight at 4˝C with a mouse monoclonal anti-metallothionein antibody (Stressgen Biotechnologies Corporation, Victoria, BC, Canada) at working dilutions of 1:100. The slides were washed several times with PBS and incubated with fluorescein isothiocyanate-conjugated γ-globulin goat anti-mouse (1:50, Sigma-Aldrich) for 30 min at room temperature in the dark. To visualize nuclei, the slides were stained with propidium iodide (a marker of nucleic acids) (1:200, Sigma-Aldrich), washed briefly with PBS and then mounted. Observations were performed by using a Leica TCS SP2 Confocal Laser Scanning Microscope (Leica Microsystems).

Mt mRNA in situ Hybridization
In situ hybridization of mt mRNA was performed according to Ebbesson and colleagues [71]. Samples of liver tissue were fixed in 4% paraformaldehyde in 0.1 M Sørensens phosphate buffer (28 mM NaH 2 PO 4 , pH 7.2) for 48 h.
After several washes in PBS (3ˆ20 min), the tissue was cryoprotect by incubation in 25% sucrose (in PBS at 4˝C overnight), embedded in Tissue-Tek O.C.T. Compound (Sakura Fintek, Zoeterwoude, the Netherlands) and stored at´80˝C. Cryosectioning was performed using a Leica CM 1850 cryostat (Leica Microsystems). Collected transversal sections of livers (12 µm) were dried at 60˝C for 10 min before processing. Using a digoxigenin (DIG)-RNA labelling mix (Roche Diagnostics, Mannheim, Germany) the digoxigenin labelled RNA probes was arranged.
The in situ hybridization probe was 540 nucleotides long. The gene was cloned using the following primers: forward 5 1 -CCC AAG CTT ATT TCT AAG GAA CTT TCA AGC-3 1 and reverse 5 1 -CCG CTC GAG TAA ATA CCA CCA TTT ATT TTA G-3 1 according to Chen and colleagues [72]. The quality and quantity of the synthesized riboprobes were assessed by agarose gel electrophoresis.
The cryostat sections were kept at room temperature for 1 h and then placed at 65˝C for 10 min. Before the rehydration in ethanol series (from 95% to 50%) the sections were washed with 2ˆSSC, incubated in 10 µg/mL of proteinase K solution in 0.1 M Tris-HCl (pH 8.0) for 3.5 min. After a post-fixation in 4% paraformaldehyde in PBS for 5 min sections were washed in PBS (2 times for 2 min), and then incubated for 3 min in triethanolamine (TEA, 0.1 M pH 8.0; Sigma-Aldrich) and acetylated for 10 min with 0.25% acetic anhydride (Sigma-Aldrich) in 0.1 M TEA.
Before dehydration in ethanol the slides were washed in 2ˆSSC (1 min). Section were then air dried for 1 h. To each section were added 200 ng of digoxigenin labelled probe in 100 µL of hybridization solution [10 mM Tris-HCl, 300 mM NaCl, 20 mM of ethylenediaminetetraacetic acid-EDTA, 0.2% tween-20, 1% blocking solution (Roche Diagnostics), 0.1% dextran sulfate (Sigma-Aldrich), 50% deionized formamide (Sigma-Aldrich)]. The sections were covered with hybri-slips (Sigma-Aldrich) and incubated in a humid chamber for 16 h at 65˝C. As control for non-specific staining a sense probe was used.
Then section were washed with RNase A (0.02 mg/mL, Sigma-Aldrich) (20 min at 37˝C and then 20 min at 65˝C). The slides were blocked with 2% blocking solution in 2ˆSSC with 0.05% Triton X-100 for 1 h and then incubated overnight in a humid chamber with alkaline phosphatase conjugated sheep anti-DIG goat antibody (1:2000, Roche Diagnostics).
The slides were washed in maleate buffer (2 times for 10 min) and finally the signal was visualized by incubation with freshly prepared visualization buffer + 100 mM Tris-HCl, 100 mM NaCl, (pH 9.5). The color reaction with chromogen substrate (3.4 µL of nitroblue tetrazolium, 3.5 µL of 5-bromo-4-chloro-3-indoylphosphate (Roche Diagnostics) in visualization buffer) was carried out for 3 h in darkness at room temperature. The reaction was terminated with stop solution (10 mM Tris-HCl, 1 mM EDTA, 150 mM NaCl, pH 8.0) and tissue was mounted with 70% glycerol. Photographs were taken using a LM LEITZ Dialux EB 20.

Quantitative RT-PCR
Extraction of total RNAs from 30 mg of liver tissue, was performed using the PureLink RNA Mini Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. To check the extraction quality, all produced RNAs were electrophoresed on a 1% agarose-formaldehyde gel. RNAs concentration was determined by spectrophotometry. For each exposure condition five replicates were performed. First-strand cDNA was synthesized from 2 µg of total RNA using the High capacity RNA to cDNA Kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's instructions. The cDNA was conserved at´20˝C until it use in a real-time PCR reaction. The gene used in our study is the metal-regulatory transcription factor 1 (mtf1, accession number NM_152981.1) and two reference genes, glyceraldehyde-3-phosphate dehydrogenase (gapdh, accession number NM_001115114.1) and actin beta 1(actb1; accession number NM_131031.1). The amplification of cDNA was checked using the TaqMan Gene Expression Assays. Real-time PCR reactions were performed in a Light Cycler (Applied Biosystems StepOn, Real-Time PCR System, Foster City, CA, USA) following the manufacturer's instructions (one cycle at 50˝C for 2 min, 95˝C for 10 min and 40 amplification cycles at 95˝C for 15 s, 60˝C for 1 min). Each 20 µL reaction contained 2 µL of reverse transcribed product template, 10 µL of master mix (TaqMan Universal Master Mix II) (Applied Biosystems), 1 µL of assay mix (TaqMan Gene Expression Assay) including a pair of unlabeled gene specific PCR primers and a TaqMan probe with a FAM dye label on the 5 1 , a minor groove binder (MGB) and non-fluorescent quencher (NFQ) on the 3 1 end) and 7 µL of H 2 O RNase-free.
The reaction specificity was determined for each reaction from the dissociation curve of the PCR product. Relative quantification of each gene expression level was normalized according to the average of the two housekeeping gene expression levels (actb1 and gapdh). Relative mRNA expression of the gene was generated using the 2´∆ Ct method [73].

Statistical Analysis
The effect of HgCl 2 exposure on MT gene expression in the liver tissue of Danio rerio was performed using Graph Pad Prism 5.00 (GraphPad Software Inc., San Diego, CA, USA) at significance level of 0.05. Assumptions of normality and homoscedasticity were tested with KS normality test and Bartlett's test, respectively. The effect of HgCl 2 exposure was compared using one-way ANOVA followed by Bonferroni's Multiple Comparison Test.