Redox and biometal status in Wistar rats after subacute exposure to fluoride and selenium counter-effects

All chemicals were of analytical grade. Sodium fluoride was purchased from Kemika (Zagreb, Croatia), sodium selenite (Na₂SeO₃x5H₂O) from Fluka (Buchs, Switzerland), perchloric acid (HCIO₄) from Fisher Scientific (Waltham, MA, USA), and nitric acid (HNO₃) and chemicals for redox status and bioelement determination from Sigma-Aldrich-Chemie (Steinheim, Germany).

The study included 20 male albino Wistar rats, obtained from the Military Medical Academy (Belgrade, Serbia). At the time of the experimental treatment, animals were eight weeks old, weighing 140–190 g. One week before the experiment, the animals were acclimatized to controlled conditions: relative humidity 60–70 %, temperature 22±2 °C, and 12 h light/12 h dark cycle. They had free access to tap water and standard food pellets, obtained from the Veterinary Institute Subotica, Serbia.

They were randomly divided into four groups of five animals. Our decision to use only five rats per group and one F^- and Se dose stemmed from the principles of 3Rs (replacement, reduction, and refinement) and animal welfare (46). The first group was control, with free access to tap water and food. The other three groups were exposed to NaF and/or Se through tap water solutions prepared on a daily basis. The fluoride group was exposed to NaF in the concentration of 150 mg/ L (F group), the selenium group to Na₂SeO₃ in the concentration of 1.5 mg/ L (Se group), and the combination group to NaF + Na₂SeO₃ in the concentrations given above (F+Se group). We selected this exposure route instead of gavage to simulate real-life exposure through drinking water as one of the main sources of exposure to F^- for the general population (47, 48). In addition, this route is less invasive for the animals. Our choice of the Se concentration was based on literature data, which indicate that 1.5 mg/ L is optimal against F^- toxicity in rats (49, 50, 51). To get an approximation of exposure, we measured water consumption for each group every day throughout the experiment, and calculated average daily water consumption per group. Based on these measurements of water consumption and body mass, we calculated that rats ingested 19.2 mg of F^- per kg of body weight (bw) a day, which is subacute exposure, considering that acute toxicity was reported at daily doses between 31 and 102 mg/ kg bw. Furthermore, our previous research (52, 53) has shown that this subacute dose can induce oxidative stress in the serum, liver, spleen, and kidney.

Experimental exposure lasted for 28 days, after which the rats were euthanised with 300 mg/kg of ketamine (100 mg/mL) combined with 30 mg/ kg of xylazinum (20 mg/mL) administered intraperitoneally (ip).

The experiment was approved by the ethics committees of the University of Belgrade School of Dental Medicine and Faculty of Pharmacy (Belgrade, Serbia; approval Nos.: 36/2 and 323-07-11822/2018-05) in accordance with the EU Directive 2010 /63/EU on the Protection of Animals Used for Scientific Purposes (54).

Body weight gain (BWG) was calculated at four time points (on days 7, 14. 21, and 28 of the experiment) using the following equation:

where cbm is the current body mass on measurement day and bbm the baseline body mass of each animal in each group.

Blood samples were collected from the euthanised animals into 3 mL heparinised syringes by direct cardiac puncture with a 21G needle and their aliquots transferred in two heparin tubes. To assess redox status parameters, we separated plasma in the first aliquot by centrifugation at 3000×g for 15 min, stored it in Eppendorf tubes (2 mL), and frozen in liquid nitrogen (-80 °C) until further analyses. The second aliquot of whole blood was stored at -20 °C until bioelement analysis.

Liver, spleen, brain, and kidney tissue samples were excised and rinsed from blood with cold saline (0.9 % NaCl). Those for redox status analysis were immediately frozen to -80 °C in liquid nitrogen, and those for bioelement and F^- analyses were stored at -20 °C.

Cu, Zn, and Fe levels were determined in 1 mL of blood and about 500 mg of organ fragments. Glass platters used for blood and organs were soaked in 10 % HNO₃ and washed with distilled water. Samples were weighed on an analytical balance (Radwag, Radom, Poland) and placed in Erlenmeyer flasks to which we added 8 mL of 69 % HNO₃ and 2 mL of 71 % HCIO₄ for digestion. Samples were then placed in a sand bath (Elektron, Trstenik, Serbia) at 200 °C to dry by evaporation. Cu, Zn, and Fe were quantified with flame atomic absorption spectrometry (FAAS, 240FS AA, Agilent Technologies Santa Clara, CA, USA) against an external standard, while calibration was performed with the ICM-100 calibration standard solution (Agilent Technologies) in the following increasing concentrations: 0.10 mg/L, 0.20 mg/L, 0.50 mg/L, 1 mg/ L, 5 mg/ L, and 10 mg/ L. For analytical accuracy we used whole blood level 2 (SeronormTM, Sero, Billingstad, Norway) and 1577c - bovine liver (LGS Standards, Teddington, UK) as standard reference materials (SRM), and the obtained recovery ranged from 88.4 % to 106.2 %. Absorbance was measured for Zn at 213.9 nm, for Cu at 324.8 nm, and for Fe at 248.3 nm.

After euthanasia, rat liver, spleen, brain, and kidney tissue samples were kept in a freezer at -20 °C and F^- determined as described elsewhere in detail (61). Briefly, each diffusion cell (Uniplast, Čačak, Serbia) was added 0.5 mL of 1 mol/ L NaOH in ethanol and placed in the laboratory oven (SAUTER SA, Basel, Switzerland) at 55 °C for 2 h for ethanol to evaporate and let a thin layer of NaOH form on the cap due. The diffusion cells were then ready to receive tissue samples (200–300 mg), 1.5 mL of 40 % AgClO₄, and 1.5 mL of 70 % HClO₄ to start microdiffusion at 55 °C for 24 h. Microdiffusion allows F^- (released thanks to 70 % HClO₄) to react with NaOH and form a thin layer of NaF on the cap. This layer was washed with 5 mL of deionised water, moved into a polyethylene dish, and mixed with 5 mL of TISAB buffer solution [57 mL glacial acetic acid, 58 g sodium chloride, 300 mg sodium citrate, and water up to 500 mL, pH 5–5.5 (pH meter Iskra MA 5735)]. Fluoride concentration (mg/kg) was determined by measuring the potential (mV) of the prepared solution using a fluoride ion-selective electrode (type 800, Consort Belgium, Brussels, Belgium) and calculating its negative logarithm, as it equals F^- concentration The electrode was checked against NaF standards in the concentrations of 0.01 mg/ L and 0.1 mg/ L (Merck Millipore, Burlington, MA, USA). Another solution with 0.05 mg/ L of NaF standard (Merck Millipore) was used for control. To ensure method accuracy the electrode was checked against control after the measurement of every 15^th sample.

Statistics were run on the GraphPad Prism 8 software (GraphPad Software, Inc., San Diego, CA, USA). The normality of distribution for variances was determined with the Shapiro-Wilk test. One-way ANOVA was used for normal distribution, followed by Fisher’s least significant difference (LSD). Data not distributed normally were run with the Kruskal–Wallis test, followed by Dunn’s post-hoc test. The level of significance was set to p<0.05.

Table 1 shows changes in body weight gain by treatment groups over the 28 days of the experiment. Significantly lower BWG was observed in the F group compared to control only on day 7 of measurement. By day 28, the two groups did not differ significantly, which is in line with some earlier reports (62, 63, 64). However, BWG seems to depend on the NaF dose, as some authors (65) reported higher BWG than ours at a lower NaF concentration of 50 mg/L NaF over 11 weeks, and others (66) reported lower BWG at NaF concentration of 500 mg/L over 60 days. Interestingly, BWG in the F+Se group was lower than control and/or the Se group throughout the experiment.

Figure 1 shows average water consumption by groups over the 28 days of our experiment. It was significantly lower in all three exposed groups compared to control. Low water consumption is the most prominent in the Se group and may have affected some of the oxidative stress parameters. Our findings confirm some earlier reports, in which Se supplementation (2 mg/L) also resulted in lower water consumption in rats (67, 68).

Figure 1

Table 2 shows Cu, Zn, and Fe concentrations after subacute exposure to NaF alone and in combination with Se. NaF alone significantly increased Cu and Zn in the blood but lowered Cu in the liver and Zn in the kidney. It also lowered Fe in the brain and kidney. No significant changes were observed in the spleen.

Fluoride effects on trace metals are still controversial and have not been explored enough. One recent study (70) reported reduced serum Cu and Zn and elevated Fe in children with chronic endemic fluorosis, which, according to authors might be related to oxidative stress. Another epidemiological study in children with endemic fluorosis (87), in turn, reported lower serum Zn but higher Cu. We, in contrast, found increased blood Cu and Zn levels. Other animal study data are also conflicting. In rats exposed to 25 mg/ L of NaF, kidney Zn levels were elevated and Cu did not change significantly (88). We too found stable kidney levels of Cu but also a drop in Zn levels. In the liver, Cu dropped, while Zn remained similar to control. Similar observations for Cu were reported by Kanwar et al. (89) after exposure to different NaF concentrations.

As for the Fe content, our findings seem to confirm the hypothesis that Zn has a role in Fe metabolism and absorption (90), as both Zn and Fe were lower in the brain and kidney tissues.

Se in combination with NaF restored some metal levels to near control but also significantly decreased liver Cu and Fe and brain and kidney Zn compared to control. Alone, it had varying effects on metal levels across the tissues (Table 2). These findings are in line with the bimodal action of Se, both beneficial and adverse, reported by other studies (91, 92, 93), which earned it the nickname of “double-edged sword”.

Table 3 compares F^- concentrations across the liver, kidney, spleen, and brain tissues. Liver had the lowest concentrations across all groups, whereas kidneys had the highest in the groups that did not receive Se (control and F) and spleen in the groups that did (Se and F+Se). High kidney levels are expected, as F^- is mainly deposited in the bone, and excreted by the kidney (94). On the other hand, the Se+F group had lower liver, brain, and kidney F^- levels than control (Figure 7), which may point to an antagonism between Se and F, except in the spleen, in which the F+Se combination resulted in the highest F^- levels.

Figure 7