In vitro Sodium Fluoride Treatment Significantly Affects Apoptosis and Proliferation in the Liver of Embryonic Chickens

,

. Acute and high-dose exposure to fluorides can also induce oxidative stress (CAO et al. 2015) and can cause renal disfunction (LANTZ et al. 1987), reproductive toxicity (GUPTA et al. 2007) and infertility (FRENI 1994). Moreover, a high dose of fluoride, due to its ability to cross the placental barrier, may directly damage a developing mammalian foetus (GUNA SHERLIN & VERMA 2001). The kidney eliminates 50-80% of the absorbed fluoride (GUAN et al. 2000), but the liver, as a site of very active metabolism, is especially susceptible to fluoride toxicity (XIONG et al. 2007). Exposure to excessive fluoride induces damage in the liver and kidneys (SHIVASHANKARA et al. 2000). It is believed that waterborne fluoride may affect the liver and kidney functions in animals and humans in a dose-dependent manner (GUAN et al. 2000).
Apoptosis, or programmed cell death, is a key process in regulating the maturation, remodelling, growth and development of a living organism. It has been detected in chickens as early as 3 h after the onset of incubation (HIRATA & HALL 2000). Apoptosis can be triggered through different intracellular signalling pathways. To maintain tissue homeostasis, many regulatory and inhibitory mechanisms must play a role in controlling apoptosis. Caspases (CASP) are a family of intracellular cysteine proteases that are linked to both the initial and final stages of apoptosis. In birds, as in other animals, the orthologous caspase mRNAs have been characterised (chicken CASP1, CASP2, CASP3, CASP6, CASP7, CASP8, CASP9, CASP10, CASP14 and CASP18) and two general pathways involving initiator caspases have been identified. One upstream pathway, which is promoted by the activation of death receptors, results in the processing of the initiator CASP8 and/or CASP10 (STRASSER & NEWTON 1999). The alternative pathway involving an initiator caspase occurs after the release of mitochondrial cytochrome C (EKERT et al. 2001). Cytosolic cytochrome C mediates the formation of the apoptosome complex, which is a highmolecular-weight complex consisting of APAF1 and CASP9. The autoactivation of either caspase-8 or caspase-9 eventually initiates the processing of an effector caspase (CASP3, CASP6 or CASP7) and the full potentiation of the caspase cascade. Caspase activation is regulated by the cellular pro-and antiapoptotic BCL2 family of proteins. In particular, anti-apoptotic members of the BCL2 family, such as BCL2, form ion channels within the mitochondrial membrane and help to maintain mitochondrial integrity. In the past few years, a novel group of cell death regulators have emerged, known as the Transmembrane BAX Inhibitor-1 Motif-containing (TMBIM) protein family. This group of proteins is composed of at least six highly-conserved members that are expressed in mammals, with homologues in insects, fish, plants, viruses and yeast. Different studies indicate that all TMBIM family members exert inhibitory activities in different settings of apoptosis. TMBIM1 affects the apoptosis signalling induced by FasL (ROJAS-RIVERA & HETZ 2015).
In our previous study, we found that an in vitro sodium fluoride (NaF) treatment significantly affected the mRNA expression and immunoexpression of oxidative stress regulatory enzymes in the embryonic gonads of chickens (GRZEGORZEWSKA et al. 2020b). NaF may also affect the expression of steroid hormone receptors in an embryonic chicken's ovaries and testis (GRZEGORZEWSKA 2020). We also demonstrated that chicken embryos may be used as an animal model in studies concerning NaF embryotoxicity. There are many reports concerning the effects of environmental factors on apoptosis in an embryonic chicken's liver. They have revealed that electromagnetic fields (LAHIJANI et al. 2011), zinc oxide nanoparticles (HAO et al. 2017) or aflatoxin b1 (AFB1) (ELWAN et al. 2021) may all induce apoptosis in the livers of chicken embryos, but the effects of NaF on apoptosis and proliferation in the embryonic liver have not been fully elucidated. The aim of this study, therefore, was to analyse the effects of an in vitro treatment of embryonic chicken livers with increasing doses of NaF on the tissue structure, number of apoptotic cells, proliferation index and the immunolocalisation of APAF1 and CASP3, as well as on the mRNA expression of apoptosis regulatory proteins.
R e s e a r c h h y p o t h e s i s: Exposure of embryonic livers from chickens to NaF will affect the tissue structure and the expression of apoptosis regulators.

Ethics approval
The chicken embryo and CAM model may be used as an alternative model and does not require the approval of a Local Ethics Committee (Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes).
The chemicals used in the tissue preparation and the Real-time PCR analysis were purchased from the following companies: StayRNA (A&A Biotechnology,

Animals and the experimental procedure
The experiment was conducted on n=60 fertilised eggs of Hy Line chickens purchased from a local breeder (Tarnów, Poland).
The eggs were incubated in a Brinsea 190 Advance automatic incubator under standard conditions (temp. 37.5 o C; humidity 55%). On the 8 th day of incubation, the quality of the eggs was controlled during an observation (candling) by ovoscope, and the unfertilised eggs and dead embryos were eliminated. On the 14 th day of embryogenesis, the embryos were sacrificed and the total livers were isolated.
In the case of chickens, the left lobe of the liver is divided into two parts (middle and lateral), while the gallbladder is located on the surface of the right lobe. Therefore, the livers were isolated very carefully, to avoid damaging the gallbladder. The livers were then dissected, immediately placed in ice-cold PBS, and rinsed twice with PBS containing an antibioticantimycotic solution. Each lobe of the liver was put into the well of 24-well plates containing 1 ml of Eagle's medium, supplemented with a 0.05% bovine serum albumin and 2 ìl antibiotic antimycotic solution, and with NaF at concentrations of 1.7, 3.5, 7.1 or 14.2 mM, or a vehicle (Eagle's medium with a NaF solvent -H 2 O). The liver lobes were incubated separately, with the left lobe used for further immunohistochemical analyses (n=6 in each group), and the right lobe used for the qPCR analyses (n=6 in each group). The livers were cultured for 6 h at 38°C in a humidified atmosphere of 95% air and 5% CO 2 . The concentration of NaF in the medium was established by considering the research data (WEI et al. 2014;FU et al. 2016) and as tested by us previously in other experiments (GRZEGORZEWSKA 2020a;GRZEGORZEWSKA et al. 2020b).
Following incubation, the tissue samples were immediately placed in StayRNA and were stored at -20 o C until total RNA isolation, or were fixed in a freshly prepared 10% (v:v) buffered formalin pH=6.9, then processed and embedded in paraffin wax for the subsequent histological analysis, DNA staining, localisation, counting of the apoptotic and proliferating cells, and the immunodetection of CASP3 and APAF1.

RNA extraction, reverse transcription and real-time PCR
The RNA isolation and reverse transcription were conducted as described previously (GRZEGORZEWSKA et al. 2020a), and the EvaGreen real-time PCR analyses were performed according to the manufacturer's instructions. Briefly, the total RNA (2 ìg from each tissue extracted with TRI-reagent) was reversetranscribed with a High-Capacity cDNA Reverse Transcription Kit including random primers. The samples were incubated in a thermocycler (Mastercycler Gradient; Eppendorf, Hamburg, Germany). in accordance with the following thermal profile: 25 o C for 10 min, 37 o C for 120 min and 85 o C for 5 min. The obtained cDNA was used in a real-time qPCR for cytochrome C (CYCS), APAF1, Transmembrane BAX Inhibitor Motif Containing 1 (TMBIM1), BCL2, caspase-3 (CASP3), caspase-9 (CASP9), and for glyceraldeyde-3-phosphate dehydrogenase (GAPDH) as a housekeeping gene. The amplifications were performed in a 96-well thermocycler (StepOnePlus; Applied Biosystems, USA), according to the following recommended cycling programme: 15 min at 95 o C, 40 cycles of 15 s at 95 o C, 20 s at 62 o C and 20 s at 72 o C. The primers (Table 1) were synthesised according to previously used and tested sequences (YIN et al. 2016;KHILLARE et al. 2018). The singleplex realtime qPCRs for the examined genes were performed in a 10 ìl volume containing: 2 ìl 5x HOT FIREPol EvaGreen qPCR Mix Plus, 0.12 ìl Primer Forward (10 pmol/ìl), 0.12 ìl Primer Reverse (10 pmol/ìl), 1 ìl cDNA (10x diluted sample after the RT template) and PCR grade H 2 O up to 10 ìl. Each sample was analysed in duplicate. A no-template control was included in each run. A relative quantification of the examined genes was calculated after a normalisation with the GAPDH transcript and by employing the expression in the liver tissue of the control chickens as the calibrator by using the 2 -ÄÄCt method (LIVAK & SCHMITTGEN 2001). A quantification was performed using StepOne integrated software.

Immunohistochemical and histological analyses H&E staining
For the histological analysis, deparaffinised and rehydrated 6-ìm-thick sections of the chicken livers were stained according to the standard haematoxylin and eosin method: staining with a Harris Haematoxylin solution for 20 minutes, differentiation in 70% ethanol containing 1% HCl, and staining with an Eosin 1% aqua solution with acetic acid (according to Harris' Hematoxylin protocol, National diagnostics).
Feulgen staining -DNA detection For the DNA detection, deparaffinised and rehydrated 6-ìm-thick sections of the chicken livers were stained according to the protocol for Feulgen staining. Acid hydrolysis was performed with a HCl solution. The next step was a treatment with Schiff's reagent and differentiation with sodium sulphite (MELLO & VIDAL 2017).

TUNEL assay
For the evaluation of apoptosis, the chicken livers were fixed in 4% formaldehyde in phosphatebuffered saline (pH 7.4) and were embedded in paraffin wax by a routine procedure. Deparaffinised and rehydrated 6-ìm-thick sections of the liver tissues were incubated with proteinase K (20 ìg ml -1 ) in 10 mM Tris-HCl, pH 7.4 at 37 o C for 20 min. Apoptotic cells were detected by TUNEL (GAVRIELI et al. 1992;GRZEGORZEWSKA et al. 2020a) using an in situ cell death detection kit, POD according to the manufacturer's instructions (Roche Diagnostics GmbH, Mann-heim, Germany). Negative controls were prepared by incubation without terminal deoxynucleotidyl transferase. To visualise the immunoreaction products, sections were incubated with a DAB and H 2 O 2 mixture for approximately 2 min. The slides were examined under an Axio Scope light microscope with an Axiocam 503 colour camera and ZEN 2.3 pro software (Carl Zeiss, Germany). The apoptotic cells (TUNEL-positive) were counted with a computerised image analysis system on 10-15 random areas of each liver tissue and were averaged for each embryo. The resulting value (the apoptotic index: AI) was shown as the number of TUNEL-positive cells per 100 cells counted. The mean value was calculated from a total of six embryos.

I m m u n o h i s t o c h e m i c a l l o c a l i s a t i o n o f C A S P 3 , A P A F 1 a n d P C N A
The liver tissue sections (6 µm thick) were deparaffinised in xylene and rehydrated by passing through an alcohol gradient, then rinsed in water and heated in a citrate buffer (pH 6.0, 75 o C, 20 min), followed by incubation with 5% (v/v) normal goat serum in TBST at room temperature for 30 min. The sections were then incubated overnight (4 o C) with specific antibodies:  CYCS -Cytochrome C, APAF1 -Apoptotic protease activating factor 1, BCL2 -B-cell lymphoma 2, CASP3 -Caspase 3, CASP9 -Caspase 9, TMBIM1-transmembrane BAX inhibitor motif containing 1, GAPDH -Glyceraldehyde 3-phosphate dehydrogenase (housekeeping gene).
TBS, incubated with the fluorescent secondary antibody DyLight 594 Anti-Rabbit IgG (H+L) (2 h, darkness) and mounted with a VECTASHIELD® Hardset™ Antifade Mounting Medium with DAPI. Negative controls were prepared by a replacement of the primary antibody with TBST. The slides were examined under an Axio Scope light microscope with an Axiocam 503 colour camera and ZEN 2.3 pro software (Carl Zeiss, Germany). The immunopositive reaction for APAF1 and CASP3 proteins was not localised in the cell nuclei, and it was estimated only as being strong, moderate, weak or very weak. The number of PCNA-positive cells was presented as the immunoreactivity, which represents the number of immunopositive cells (

Statistical analysis
The gene expression data represents the mean relative quantity (RQ) from n=6 samples in each group. The analysis of the mRNA expression was performed in duplicate and standardised to control the expression in the liver (RQ = 1) using the 2 -ÄÄCt method (LIVAK & SCHMITTGEN 2001). The data was statisti-cally analysed by one-way ANOVA (effect of treatment) followed by Duncan's multiple range test. Log transformations were performed as needed, to maintain the homogeneity of variance and normality. The values were expressed as means ± SEM and were considered significantly different at p<0.05. The calculations were performed using SigmaPlot (Systat Software Inc., San Jose, CA, USA).

Data and model availability statement
No data was deposited in an official repository. The data is available from the authors upon request.

mRNA expression of the apoptosis regulators CYCS, APAF1, TMBIM1, BCL2, CASP3 and CASP9
The expression of mRNA encoding CYCS increased significantly in the group exposed to 1.7 mM of NaF (D1), but in the other experimental groups, the levels of CYCS were not significantly different from those in the control group (Fig. 1A). The mRNA expression of APAF1 was significantly higher in the D1, D3 and D4 NaF-treated groups (Fig. 1B). The TMBIM1 mRNA expression in the embryonic chicken livers was not affected by NaF (Fig. 1C), whereas the antiapoptotic BCL2 protein mRNA in- Fig. 1. mRNA expression of genes encoding the apoptosis regulators CYCS, APAF1, BCL2, CASP3, CASP9 and TMBIM1, with GAPDH as a reference gene, in the liver tissues of chicken embryos in the control group and following an in vitro treatment with 1.7 (D1), 3.5 (D2), 7.1 (D3) and 14.2 mM (D4) NaF. Each value represents the mean ± SEM from n = 6 (six embryonic livers). The gene expression data represents the mean relative quantity (RQ) standardised to control the expression in the liver (RQ = 1). Values marked with different letters differ significantly (p<0.05).
creased significantly in the D1 and D4 groups (Fig. 1  D). The expression of both analysed caspases (CASP3 and CASP9) mRNA was significantly higher in the D1 group, as well as that of CASP9 in the D3 group (Fig. 1E, F).

Histological analysis (Figs 2, 3)
In the NaF-treated livers stained with H&E, we observed a significant increase in the number of intracytoplasmic vacuoles and pyknotic nuclei, mainly in the experimental group where the highest dose of NaF was used (Fig. 2). Feulgen staining showed a dramatic increase in the number of early-apoptotic nuclei and apoptotic bodies in the group that was exposed to the highest NaF dose (D4) (Fig. 3).

Localisation of apoptotic cells
The number of apoptotic cells gradually increased in the NaF-treated groups, ranging from 18% in the control group to the highest value of 49% in the D4 group (Fig. 4). The highest numbers of TUNELpositive cells were observed near the blood vessels. Perivascular cells, connective tissue cells as well as hepatocytes in the developing liver also showed a significantly higher signal in the D4 experimental group.

Immunoexpression of CASP3 and APAF1
The expression of apoptosis regulatory factors increased as follows: CASP3 immunopositivity in the D3 and D4 groups was significantly more intensive in single cells in the embryonic liver (Fig. 5), whereas that of APAF1 in the D3 and D4 groups increased in the perivascular cells, and hepatocytes were localised mainly near the portal vein as well as the blood vessel walls (Fig. 6).

Localisation of proliferating cells (PCNA antigen)
The proliferating cell number decreased significantly in a dose-dependent manner, ranging from 84% in the control group to 53% in the D1 group, 35% in the D2 group, 19% in the D3 group and 5.5% in the D4 group (Fig. 7).

Discussion
In the present study, changes in the mRNA expression as well as in the immunolocalisation of apoptosis regulators after a treatment with NaF were observed in the livers of chicken embryos. In this experiment, four doses of NaF, ranging from 1.7 to 14.2 mM, were used (75-600 ìg NaF/ml, i.e. per well). We did not observe a dose-dependent effect of the NaF-treatment on the mRNA expression of apoptosis regulators. However, this may have been a result of a very short time for the embryonic livers incubation. Phenotypic variability may also affect the adaptogenes expres- sion. The lowest D1 dose significantly affected the mRNA encoding proteins regulating the intrinsic pathway of apoptosis activation (CYCS, APAF1, CASP3 and CASP9), but the mRNA expression of the antiapoptotic protein BCL2 also increased in this group. Higher doses of NaF altered the mRNA expression of APAF1 and CASP9. These results suggest that the lowest dose, 1.7 mM NaF, may affect liver cell apoptosis at the mRNA level. The mechanism of the NaF action may be associated with the mitochondrial pathway of apoptosis activation. A significant increase in the apoptotic index, with a clearly visible decrease in the number of proliferating cells, was also observed. Previously, LU et al. (2017) showed that NaF may cause oxidative stress and apoptosis in the mouse liver. They observed histopathological lesions including: hepatocellular granular degeneration; vacuolar degeneration; and necrosis with karyorrhexis, karyolysis and hypochromatosis. Similar effects we observed in NaF-treated livers of embryonic chickens. They also detected a significant increase in the mRNA expression of CASP3 after 21 days, only in the group that was exposed to the highest NaF dose (48 mg/kg b.w.), and after 42 days in those groups exposed to doses of 12, 24 and 48 mg/kg. They observed more significant effects at the protein expression level (LU et al. 2017). In the present study, a significant increase in the mRNA expression of CASP3 was observed only in the group that was exposed to the D1 NaF dose (1.7 mM). It is possible, in our experimental model, that this particle affects the protein rather than the mRNA expression after an incubation period of only 6 hours. In our study, significant increases in the apoptotic index and intensity of CASP3 and APAF1 immunopositivity, mainly in cells localised near the blood vessels, were observed in the groups exposed to the higher doses of NaF: 7.1 and 14.2 mM. In general, CASP3 is a key and common protease in both mitochondria-and death receptor-dependent pathways (EARNSHAW et al. 1999). The results of a study by LEE et al. (2008) showed that in human gingival fibroblasts, NaF increased the number of TUNEL-positive cells and induced apoptosis, with a concomitant chro-matin condensation and DNA fragmentation. NaF has also a significant effect on MAP kinases, leading to the arrest of cell division, and as a result it may induce apoptosis processes in odontoblast-like cells (KARUBE et al. 2009). The cell viability after a treatment with 5-40 mM of NaF for 12 h decreased from 100% to 11%. Approximately 60% of the cells survived an exposure to 20 mM NaF for 6 h. In addition, NaF increased the level of cytochrome c released from the mitochondria into the cytosol and enhanced the activities of CASP9, CASP8 and CASP3. An upregulation of Fas-ligand, a ligand of the death receptor in the NaF-treated group was also detected. BCL2, a member of the antiapoptotic BCL2 family, was downregulated, whereas the expression of BAX, a member of the proapoptotic BCL2 family, was unaffected in the NaF-treated cells (LEE et al. 2008). In the embryonic chicken liver, we did not detect any differences in the mRNA expression of TMBIM1, which is an inhibitory apoptosis regulator in the receptor-dependent apoptosis induction pathway. This suggests that the intrinsic pathway of a programmed cell death is the main pathway affected by NaF in the liver of an embryonic chicken.
The mechanism of fluorine toxicity is very intricate. The high electronegativity of the F ion allows it to interact with both the organic and inorganic compounds of cells. Furthermore, fluorine is an anion with strong penetrability (ZHANG et al. 2016). It has a great affinity for the Ca 2+ , Mg 2+ and phosphate present in the culture medium and within cells, so the active agent could interfere with DNA polymerase and RNAase activities. The fluoride anion also forms a strong hydrogen bond with purine and pyrimidine bases, and may thus disrupt the structures of both DNA and RNA, and may interfere with their synthesis or enhance the frequency of base-pair errors during DNA replication. Fluoride has been found to inhibit protein synthesis and cell cycle progression (AARDEMA et al. 1989), and may also induce oxidative stress, which increases lymphocyte apoptosis ).
In our study, a significant decrease in the proliferating index, accompanied by significant increases in the number of intracytoplasmic vacuoles and pyknotic nuclei, as well as a dramatic increase in the number of apoptotic nuclei and apoptotic bodies, was demonstrated in the group exposed to the highest NaF dose (D4). Previously, in a study with human osteosarcoma MG-63 cells as a model, it was shown that 5 mM of NaF may stimulate proliferation, whereas 20 mM of NaF inhibits cell proliferation and induces apoptosis (WEI et al. 2014). The results of an in vitro study by WANG et al. (2004) revealed that in normal human primary hepatocytes incubated with NaF (80 µg/ml) for 12 h, the number of cells significantly increased in the S phase, but the treatment had no effect on the cell numbers in the G0/G phase and G2/M phase. This finding indicates that fluoride can disturb signal transduction in the cell cycle and cause cell arrest in the S phase. HE & CHEN (2006) analysed the effect of fluoride, at a dose of 150 mg/l in drinking water for 4 weeks, on the cell cycle and proliferation index of oral mucosal cells and hepatocytes in rat specimens. Both the number of hepatocytes in the G0/G1, S and G2/M phases and the number of proliferating cells decreased. However, the apoptotic index increased sig- nificantly (approximately 2-fold) in the NaF-treated group (HE & CHEN 2006). Many studies have demonstrated that NaF may affect embryonic development. Previously, FU et al. (2016) showed that 1 mM of NaF did not significantly affect the proliferation of human embryonic stem cells, but disturbed the gene expression patterns of hESCs during embryonic differentiation. Higher doses of NaF (>2 mM) markedly decreased the viability and proliferation of these cells (FU et al. 2016), whereas the treatment of mouse embryonic stem cells with NaF doses greater than 1 mM reduced their viability and DNA synthesis, and induced cell cycle arrest in the G(2)/M phase as well as cell death, mainly by apoptosis rather than necrosis (NGUYEN NGOC et al. 2012). BAI et al. (2010) analysed the effect of fluoride on the renal cell population of chickens. They found that a fluoride treatment greatly increased the population of these cells in the early stage of apoptosis, and at higher doses, also increased the population of renal cells in end-stage apoptosis. These results suggest that a high dietary F intake (400-1200 mg F/kg) causes G0/G1 arrest and induces apoptosis of the renal cells in chickens, which likely influences the excretion and retention of fluoride in the body.
Our results reveal that NaF, by changing the balance between apoptosis and the proliferation of hepatocytes, may disturb the development and function of the liver in embryonic chickens. The World Health Organization (WHO) states that the reference range for fluoride in drinking water is 0.5-1.5 mg/l (FAWELL et al. 2006). The American Environmental Protection Association (EPA) and the American Public Health Service (PHS) previously reported the level of fluoride in drinking water to be 0.7-1.2 mg/l, while the PHS revised its optimal value to 0.7 mg/l (USA, EPA Regulations). The EPA states that there is no need for a reference update for the time being, but it will continue to follow up on the research conducted on this subject and refer to the reference regulation when necessary (KIRMIT et al. 2020). The risk of exposure to NaF should be considered when determining the standards for human and animal exposure to this compound, and the chicken embryo may be used as an animal model to analyse the mechanisms of toxicity of NaF. Most of the literature data shows that the toxicity of fluoride is extremely dangerous for developing organisms. It should be emphasised that the chicken embryo is included in the inexpensive, generally available, alternative research models, and that research with the use of chicken embryos is recommended and complies with the guidelines of ethical committees. The individual development stages of a domestic chicken embryo are very well known and described, which is an additional advantage of this model.