Manganese itself preferentially accumulates in the cellular mitochondria (Gunter 2006; Sarkar et al. 2018) through the mitochondrial Ca2 + uniporter, where it binds to the mitochondrial membrane or matrix proteins (Gavin et al. 1999). Due to this preference, Mn can disrupt mitochondrial function by three main mechanisms: inhibiting energy transduction, inducing mutations in the mitochondrial genome, and through the improved generation of free radicals. All forms are interconnected and possibly result in metabolic dysfunction (Brown and Taylor 1999).
Mitochondria is involved in the intrinsic pathway of cell apoptosis (the path by which we believe to justify the findings of our study) due to the activity of releasing cytochrome c from the intermembrane space to initiate the activation of caspases in the cell cytosol (Wang and Youle 2009). In vertebrates, the release of these proteins is a consequence of compromising the integrity of the permeabilization of the outer mitochondrial membrane, which acts controlling the pro-apoptotic members of the Bcl-2 family (Martinou and Youle 2011). The permeabilization of the outer mitochondrial membrane can occur through an increase in the reactive oxygen species (ROS) production in the mitochondria. It contributes to the permeabilization of the inner mitochondrial membrane and triggers a series of factors that will consequently result in the complete loss of the mitochondrial function (Solá et al. 2013).
In the present study, we detected significant differences in the percentage of apoptotic cells stained with TUNEL in the zebrafish telencephalon exposed to MnCl2 (at 4.0 mg L-1 for 72 hours), and the tendency to increase TUNEL positive cells over the 21 days of exposure and recovery. This fact coincides with the lower rate of mitochondrial activity detected in the animals' telencephalon (4.0 mg L-1 in the 21st day), which may be associated with the apoptotic intrinsic pathway activation. This mitochondrial function oscillation can promote the complete mitochondrial dysfunction of most telencephalic cells, generating the release of cytochrome c in the cell cytosol and resulting in the caspase and apoptosis activation, which reflects the dataset obtained in 21 days of exposure.
Countless authors have already associated the increased production of ROS as the main mechanism of action involved in the neurotoxicity caused by manganese exposure (Milatovic et al. 2007; Lebda et al. 2012), which can be mainly explained by preferential metal accumulation by cellular mitochondria (Gunter et al. 2006). Green and Reed (1998) and Martinez-Finley et al. (2013) reinforce our idea that ROS generation and the subsequent decrease in the mitochondrial membrane potential affect the mitochondrial permeability transition, leading to organelle swelling, rupture of the outer membrane, and the release of various apoptogenic factors in the cytosol.
It is noteworthy that after 30 days of recovery in decontaminated water, animals showed an increase in the mitochondrial activity in relation to the observed in 21 days. It may be associated with the observation of new cells due to the rapid neurogenesis occurring in the zebrafish (Ghosh and Hui 2016; Ogai et al. 2012), which can be triggered by the secondary role of cell proliferation that some caspases precisely play to compensate cell loss (Pérez-Garijo 2018). At the same time, we still observe the tendency to increase the rate of TUNEL positive cells, indicating that the increase in cell apoptosis can be persistent after a short period of recovery, even considering the great capacity of neural regeneration of the species.
Although such changes have been observed at the cellular level, they have not resulted in major behavioral changes, corroborating with a previous study performed in our laboratory using the same concentrations of MnCl2 (Rodrigues et al. 2020). The low incidence of behavioral changes reported in both studies can be justified by the fact that low concentrations of MnCl2 are not sufficiently capable of damaging motor neurons at a rate (greater than the natural rate) of neurogenesis in the zebrafish telencephalon. This occurs throughout the dorsal and ventral telencephalon in the zebrafish (Adolf et al. 2006), considering that proliferative periventricular zones are not only presented internally but also in the outer surface of the telencephalon (Wullimann and Mueller 2004). Locomotor changes have been reported in zebrafish exposed to manganese but in much higher concentrations, which do not represent environmental conditions (Altenhofen et al. 2017).
In our study, there was a tendency, especially at 14 and 21 days, for animals from the 4.0 mg L-1 group to explore a greater distance than the control at the top of the aquarium. It can indicate that MnCl2 may have acted as an anxiolytic, since it is expected that the contact with the new (tank test) could induce an anxious behavior in the animals. This behavior results in their permanence at the bottom of the aquarium for a longer time, carefully exploring the top, due to the greater risk of predation (Richendrfer et al. 2012). The same pattern was observed in adult zebrafish exposed for 96 hours at 63 and 126 mg L-1 of MnCl2 (Altenhofen et al. 2017). However, this behavioral parameter appears to be normalized after the recovery period in our study. Other authors reported the locomotor recovery and exploratory functions in adult zebrafish submitted to hypoxia conditions after 3 hours of exposure in normoxic conditions (Braga et al. 2013). Manuel and collaborators (2015) also reported the reduction of anxious behavior after a recovery period associated with environmental enrichment. Therefore, we can conclude that behavioral changes are liable to partial and/or total reversion when facing a decontaminated or normalized environment within the specific standards of the species, even if the cellular level changes are permanent and irreversible.
Among these persistent damages, we can highlight the intestinal histopathological changes observed in our study. The percentage of villi with leukocyte infiltration and increased epithelial thickness remained higher in animals exposed to both concentrations of MnCl2 compared to the animals in the control group in all sampling periods of the experiment, including after the recovery period. Both histological findings are common in the intestinal inflammation described for the species (Brugman et al. 2009; Zhao and Pack 2007). The increase in intestinal villi may have occurred as a manner to adapt to the fish exposure environment in order to reduce the entry of MnCl2 from the intestinal lumen into the circulation (Abraham and Cho 2009). It may even result in milder damage in the other organs, since the metal ingestion by fish can also occur through the ingestion of contaminated food and particulate matter, besides through ion exchange, gills, and skin (Cáceres-Vélez et al. 2019).
We emphasize that the increase in intestinal cell infiltration, i.e., the presence of inflammation in animals exposed to 4.0 mg L-1 of MnCl2, occurred in the same experimental periods in which we observed a better exploration behavior in the upper zone of the aquarium (14 and 21 days, although no statistical difference was detected in 21 days). Some authors have already speculated the influence of the gut on zebrafish behavior, especially regarding the issue of intestinal microbiota (Davis et al. 2019), which is essential for normal stress responsiveness and anxiety-related behavior (Luczynski et al. 2016).
Although we did not make such analysis in our study, it has shown that manganese reduces the richness of intestinal bacteria (Wang et al. 2020a), resulting in the intestinal and nervous system inflammation through the entry of metal into the bloodstream and hematoencephalic barrier (Wang et al. 2020b). It can also result in neurodegeneration by promoting amyloid formation or increasing inflammatory responses to endogenous neuronal amyloid (Friedland and Chapman 2017), which has also been associated with MnCl2 exposure (Wang et al. 2020a). Therefore, the intestinal inflammation and/or the possible alteration in the intestinal microbiota derived from the MnCl2 exposure may have modulated the locomotor behavior of our animals, as it had already occurred in another study carried out with cadmium, for example (Xia et al. 2020).
When absorbed by the intestine, manganese enters the bloodstream and is transported to the liver through the portal vein (Barceloux 1999). Regardless the absorption pathway, the liver is one of the tissues with greater mitochondrial activity leading to manganese accumulation (Williams et al. 2012). Thus, this organ has proved to be useful in the histopathological evaluation of animals.
Some authors believe that histopathological damages in animals’ organs exposed to some pollutants may be due to biochemical damages (Macêdo et al. 2020). Hepatocyte vacuolization, for example, is associated with the protein synthesis inhibition, depletion of energy, microtubules disintegration, or changes in the substrate use (Hinton and Lauren 1990). Studies have shown that fish hepatocytes tend to be more vacuolated than mammalian hepatocytes, especially in animals kept in captivity, due to artificial diet and housing conditions (Ferguson 1989; Gingerich 1982).
Besides this adversity, animals treated with both concentrations of MnCl2 demonstrated a high rate of cell vacuolization compared to the control animals, indicating a higher glycogen or lipid content in their hepatocytes. It may evolve to cell degeneration when the content (glycogen or lipid) accumulates to the point of distention of the hepatocyte cytoplasm, resulting in cytomegaly (Wolf and Wheeler 2018). In our study, after the recovery period, animals exposed to 0.5 mg L-1 demonstrated an improvement in the recovery rate, even compared to the 72-hour group. However, the specimens of the concentration of 4.0 mg L-1 remained with a very high hepatocyte vacuolization rate.
This response pattern difference may be associated with the relatively smaller capacity of fish (in relation to mammals) to metabolize xenobiotic substances, by eliminating such unchanged substances through their gills (Parkinson 2008; Wolf and Wolf 2005). Therefore, we believe that at the lowest exposure concentration, fishes in our study may have had the ability to not completely metabolize the metal available in the water. The ability to recover/neutralize the rate of vacuolated hepatocytes in animals exposed to 0.5 mg L-1 and the degeneration in 4.0 mg L-1 specimens may have been triggered by the activity of antioxidant defense enzymes, such as superoxide dismutase (Wang et al. 2014).
The same pattern of response was observed in the behavior of the studied offspring, as the main behavioral changes were observed at 5 dpf and 7 dpf. It indicates that the damage to the locomotor system also does not appear to be persistent, which can be justified by the absence of the apoptosis increase in the encephalic region of the larvae evaluated with AO. Although there was a generational transposition of some damages, these alterations do not persist after 10 days of life of these larvae, indicating that the next generations of adults from parents exposed to low concentrations of this pollutant may not undergo major changes in their biology.
As we look back at the initial aim of our study, the monitoring over different experimental periods allowed us to identify that the observed effects, especially in 4.0 mg L-1 of MnCl2, began to be expressed by the animals after a subchronic exposure of 14 days, which reflected in behavioral, intestinal, and mitochondrial changes, whereas the damage in the liver was observed after 72 hours of exposure.
Some of these damages were reversed after the recovery period, for example, anxiolytic behavior and intestinal damage. However, cell vacuolization in hepatocytes remained too high in the 4.0 mg L-1 group even after recovery. This liver overload, which remained even after recovery, could activate other biochemical and morphological changes in the animals over time. The chronic exposure of adult animals to MnCl2 resulted in some small changes in the larval behavior of the offspring, which were attenuated during development. Therefore, according to the parameters evaluated, we can conclude that the generational transposition of the damage caused by these environmentally relevant concentrations of MnCl2 is not very active, at least in these exposure times.