A possible mechanism for combined arsenic and fluoride induced cellular and DNA damage in mice

Abstract

Arsenic and fluoride are major contaminants of drinking water. Mechanisms of toxicity following individual exposure to arsenic or fluoride are well known. However, it is not explicit how combined exposure to arsenic and fluoride leads to cellular and/or DNA damage. The present study was planned to assess (i) oxidative stress during combined chronic exposure to arsenic and fluoride in drinking water, (ii) correlation of oxidative stress with cellular and DNA damage and (iii) mechanism of cellular damage using IR spectroscopy. Mice were exposed to arsenic and fluoride (50 ppm) either individually or in combination for 28 weeks. Arsenic or fluoride exposure individually led to a significant increase in reactive oxygen species (ROS) generation and associated oxidative stress in blood, liver and brain. Individual exposure to the two toxicants showed significant depletion of blood glutathione (GSH) and glucose 6-phosphate dehydrogenase (G6PD) activity, and single-stranded DNA damage using a comet assay in lymphocytes. We also observed an increase in the activity of ATPase, thiobarbituric acid reactive substance (TBARS) and a decreased, reduced and oxidized glutathione (GSH ∶ GSSG) ratio in the liver and brain. Antioxidant enzymes like superoxide dismutase (SOD), catalase and glutathione peroxidase (GPx) were decreased and increased in liver and brain respectively. The changes were more pronounced in liver compared to brain suggesting liver to be more susceptible to the toxic effects of arsenic and fluoride. Interestingly, combined exposure to arsenic and fluoride resulted in less pronounced toxic effects compared to their individual effects based on biochemical variables, IR spectra, DNA damage (TUNEL and comet assays) and histopathological observations. IR spectra suggested that arsenic or fluoride perturbs the strength of protein and amide groups; however, the shifts in peaks were not pronounced during combined exposure. These results thus highlight the role of arsenic- or fluoride-induced oxidative stress, DNA damage and protein interaction as the major determinants of toxicity, along with the differential toxic effects during arsenic–fluoride interaction during co-exposure. The study further corroborates our earlier observations that at the higher concentration co-exposures to these toxicants do not elicit synergistic toxicity.

---

Introduction

Arsenic and fluoride are recognized worldwide as some of the serious inorganic contaminants of drinking water. Many studies have reported as regards to simple fluorosis and arsenicosis, but the knowledge of the joint action of these two elements is lacking and the results derived from previous studies are inconclusive. Individual exposure to arsenic or fluoride is known to exert their toxicity through oxidative stress by generating reactive oxygen species.^1,2 Free radical generation has been reported in cells treated with arsenite or fluoride.^3,4 The increased oxidative stress at the cellular level is proposed to be an important mediating factor in the detrimental effects of chronic arsenic or fluoride toxicity.⁵

Arsenic exposure has been shown to induce a significant increase in the formation of hydroxyl radicals in soft tissues.⁶ Reactive Oxygen Species (ROS) have been demonstrated to be present in rat brain exposed to arsenic.^7,8 In addition, arsenic is known to decrease glutathione^9,10 antioxidant enzyme activity, but increases the oxidant production.¹¹ Arsenic induced free radicals affect a wide array of cellular macromolecules like lipids, proteins and DNA,^3,12 which ultimately leads to cell death. Arsenic compounds are known to induce gene amplification, cellular transformation, DNA cross-links and DNA strand breaks in animal and human cells.¹³ Genotoxic effects caused by arsenic are implicated in carcinogenic outcomes. The proposed mechanisms of arsenic induced carcinogenicity are interference with DNA repair processes and generation of reactive oxygen species.¹⁴ Base modifications are known to be potential biomarkers of oxidative stress.¹⁵ Wei et al. (2002) and Liu et al. (2002) reported that inorganic arsenite induced tissue damage might be attributed to necrosis.^16,17 On the other hand, studies have shown arsenic to induce apoptosis in tumor cell lines.^18,19

The genotoxic effects of fluoride on different tissues have not been extensively investigated. The current information about genotoxic potential of this is contradictory, older and inconclusive.²⁰ According to some authors, fluoride does not induce DNA damage.^21,22 However, some authors have observed the genotoxic potential of fluoride in rats and human cells.^23,24 The impact of fluoride on free radical generation in soft tissues has been investigated by many authors^25,26 however, the effect of fluoride on DNA damage in lymphocytes and its possible relation with oxidative stress requires further research exploration.

Although, there are few reports suggesting effects of individual exposure to arsenic or fluoride on DNA damage and cellular deformities however, relatively little is known about effects of their combined exposure on structure and metabolism of various tissues. It has been reported that fluoride and arsenic interfere with the structure and functions of liver of different animal models of both sexes.^27,28 The effects of chronic arsenic poisoning in humans cause liver lesions, portal hypertension, infiltration of inflammatory cells in the periportal area, and elevated levels of serum amino transferases which are indicative of hepatocellular necrosis.²⁹ Recently, the occurrence of hepatotoxicity in rabbits and calves by fluoride has also been reported.³⁰ Shah and Chinoy (2004) studied effects of co-administration of arsenic and fluoride on the physiology and histology of brain and reported vacuolated regions with lymphocyte infiltration. However, the mechanism of these effects is still unclear.³¹

The alkaline single cell gel electrophoresis assay, also known as the comet assay, permits evaluation of the genetic material integrity, detecting from DNA primary lesions and repair in mammalian cells to the discrimination between necrotic and apoptotic cells.^32,33 In the present study, we thus used the comet assay to evaluate DNA damage in lymphocytes of arsenic and/or fluoride exposed animals and evaluated markers of oxidative stress in blood and soft tissues to find a possible relation of DNA damage under the condition of oxidative stress. In order to have a mechanistic approach on cellular and DNA damage we noted histological changes in liver and brain and recorded FT-IR spectra of lyophilized liver and brain tissues of animals.

Thus, the present study was planned to evaluate toxic effects of individual and combined chronic exposure to arsenic and fluoride with a view to investigate the underlying mechanisms such as protein interaction and oxidative stress and their possible correlation with arsenic–fluoride induced genotoxicity.

Materials and methods

Animal preparation and study design

All experiments were performed on healthy, adult male mice of Swiss strain, weighing approximately 30 g. Animals were obtained from Defence Research and Development Establishment (DRDE) animal facility and, prior to use, were acclimatized for 7 days in a 12 h light/dark cycle. The Animal Ethical Committee of DRDE, Gwalior, India, approved the protocols for the experiments. The animals were housed in stainless steel cages in an air-conditioned room with temperature maintained at 23 ± 2 °C. Mice were allowed to feed on a standard mouse chow diet (Ashirwad Feeds, Chandigarh, India; metal content of diet in ppm dry weight: Cu 10.0, Zn 45.0, Mn 55.0, Co 5.0, Fe 75.0) and water ad libitum throughout the experiment. Animals were randomized into 4 groups with twelve mice per group and treated as below for twenty-eight weeks:

Gr. I: normal

Gr. II: arsenic as sodium meta-arsenite (50 mg l⁻¹ in drinking water)

Gr. III: fluoride as sodium fluoride (50 mg l⁻¹ in drinking water)

Gr. IV: arsenic (same as in group II) + fluoride (same as in group III)

After twenty-eight weeks, exposure was stopped, and all mice from each group were sacrificed under light ether anesthesia, 48 h, after the last dosing. Blood was collected in heparinized vials. Brain and liver were removed, washed and perfused with normal saline to remove residual blood. All the extraneous materials were removed before weighing. The brain and liver were kept under ice-cold conditions at all times.

Institutional Animal Ethics Committee (IAEC) of Defence Research and Development Establishment, Gwalior, India, approved the protocols for the experiments.

Biochemical assay

Blood GSH concentration was determined by the method of Ellman (1959),³⁴ modified by Jollow et al. (1974).³⁵ The amount of ROS in blood and tissues was measured using 2′,7′-dichlorofluorescin diacetate (DCFDA) by the method of Socci et al. (1999).³⁶ White blood cells (WBC), red blood cells (RBC), hemoglobin (Hb), hematocrit (HCT), mean cell volume (MCV), mean cell hemoglobin (MCH), mean cell hemoglobin concentration (MCHC) and platelet (PLT) counts were measured on a Sysmex Hematology Analyzer (model K4500). G6PD activity in blood was measured by polyacrylamide gel electrophoresis (PAGE) employing the method of Ornstein (1964)³⁷ and stained as described by Scholl and Anders (1973).³⁸ Single cell gel electrophoresis or comet assay was done as per the protocol of³⁹ Singh et al. (1988), with some modifications by Rao et al. (1999).⁴⁰ This is a microelectrophoretic technique for the direct visualization of DNA damage in individual cells. A small number of irradiated cells suspended in a thin agarose gel on a microscopic slide were lysed, electrophoresed and stained with a fluorescent DNA binding dye. The electric current pulled the charged DNA from the nucleus such that relaxed and broken DNA fragments migrated further. The resulting images, which were subsequently named for their appearance as ‘comets’, were measured to determine the extent of DNA damage. Briefly, 20 μl of whole blood was mixed with 200 μl of 0.75% low melting agarose in PBS at 37 °C and quickly put onto a microscopic slide, which already had a dried layer of 0.1% agarose. A cover glass was put slowly on the slide to make a uniform layer of the agarose–cell mixture. This slide was put on ice for 5 min to allow rapid solidification of gel. Then the cover glass was removed and 200 μl of agarose was placed on the slide to make an additional layer of agarose above the layer containing the cells. The slide was again cooled for 5 min before the cover glass was removed and slides were immersed into cold lysing solution (2.5 M NaCl, 100 mM EDTA, 10 mM Tris), 1% Triton X-100 and 10% DMSA were added freshly. After 12 h of lysis, slides were transferred to a horizontal electrophoresis unit containing running buffer (30 mM NaOH and 1 mM EDTA pH 13.0). The slides were kept for 20 minutes to allow the DNA to unwind. Then electrophoresis was conducted for 20 min at 0.4 V cm⁻¹. The slides were carefully removed from the unit and flooded with neutralizing buffer (0.4 M Tris, pH 7.5). DNA staining was done with ethidium bromide (200 ng ml⁻¹) and comets were visualized in a LEICA GMBH, Germany, fluorescent microscope.

GSH and GSSG contents in liver and brain were assessed according to the method described by Hissin and Hilf (1976).⁴¹ Tissue lipid per oxidation was measured by the method of Ohkawa et al. (1979).⁴² Liver and brain SOD activity was assayed by the method of Kakkar et al. (1984)⁴³ Catalase activity was assayed following the procedure of Aebi (1984).⁴⁴ Glutathione peroxidase activity was measured by the procedure of Flohe and Gunzler (1984).⁴⁵ ATPase activity was assayed according to the procedure of Griffiths and Houghton (1974).⁴⁶ FT-IR spectra were recorded with a Perkin Elmer FT-IR Spectrometer, model BXII. Tissue pellet samples were scanned at room temperature (25 ± 1 °C) in the 4000–400 cm⁻¹ spectral range.

Histopathological examination

Liver and brain tissues were dissected out and fixed in 10% formalin–saline solution. Tissues were processed in ascending concentrations of ethanol, cleared in toluene using an automatic tissue processor (Leica TP1020), further embedded in paraffin (Leica embedding station, EG1160) and serial sections (4–5 μm thickness for liver and 12 μm for brain) were cut with a Microm HM 360 microtome. Sections were stained by the routinely used Hematoxylin and Eosin (H & E) method using standardized programme in Leica autostainer XL. Oil red O staining for fat was also carried out on formalin fixed cryosections of liver tissues.

TUNEL assay

In situ apoptosis detection by TUNEL staining (Dead End™ Fluorometric TUNEL System; Catalog No. G3250, Promega, Madison, USA) was performed using manufacturers protocol. TUNEL positive cells were identified as green fluorescence of fluorescein-12-dUTP and red fluorescence of propidium iodide in liver sections under a Leica DMLB microscope at >620 nm. A liver section pretreated with HCl was taken as positive control for confirmation of TUNEL positive cells and DNA damage.

Metal estimation

Arsenic was estimated using a Hydride Vapour Generation System (Perkin Elmer model MHS-10) fitted with an atomic absorption spectrophotometer (AAS, Perkin Elmer model AAnalyst 100). Fluoride was measured in the digested blood and tissue samples using a Fluoride ion meter (Eutech Instruments).

Statistical analysis

Data are expressed as means ± SEM. Data comparisons were carried out using one-way analysis of variance followed by Bonferroni's test to compare means between the different experimental groups. Differences between experimental groups with a p value <0.05 were considered significant. Based on the affected area in the microscopic field histopathological and TUNEL variables were analyzed on an arbitrary scale of 0 to 4 [0—nil (absent); 1—minimal (1–10% area); 2—mild (11–25% area); 3—moderate (26–50% area); 4—severe (50% < area)] and the results are presented in the form of a summary incidence table.

Results

Effect on clinical, hematological and blood oxidative stress variables

Table 1 shows the effect of arsenic and fluoride co-exposure on some haematological variables in mice. A significant decrease in RBC, Hb and HCT levels was observed in animals exposed to arsenic accompanied by a significant increase in WBC and PLT counts. On the other hand, exposure to fluoride caused a significant decrease in RBC, Hb and HCT and an increase in WBC and PLT counts. Co-administration of arsenic and fluoride did not alter WBC and Hb, while RBC and HCT levels decreased significantly compared to normal. Other hematological variables (MCV, MCH, MCHC) remained unchanged (data not shown). Exposure to arsenic and fluoride caused a significant increase in the blood ROS level (Fig. 1A) accompanied by a decreased GSH level (Fig. 1B). On the other hand, combined exposure to arsenic and fluoride significantly decreased GSH compared to normal animals however, no change in the ROS level was noted.

Table 1 Effect of individual and combined exposure to arsenic and fluoride on hematological variables in mice

	Normal	As	F	As + F
Abbreviation used and units: RBC—red blood cells as ×10^3 μl^−1; WBC—white blood cells as ×10^6 μl^−1; HGB—hemoglobin as g dl^−1; HCT—hematocrit as %; PLT—platelet as ×10^3 μl^−1. Values are mean ± SE; n = 5.^*, †, ‡Differences between values with matching symbol notations within each column are not statistically significant at 5% level of probability.
WBC	6.64 ± 0.14^*	8.90 ± 0.15^†	11.1 ± 0.33^†	8.43 ± 0.80^*
RBC	9.14 ± 0.15^*	7.36 ± 0.10^†	7.97 ± 0.13^†	7.08 ± 0.09^†
HGB	13.64 ± 0.19^*	11.02 ± 0.13^†	11.22 ± 0.29^†	12.27 ± 0.28^*
HCT	46.2 ± 0.19^*	37.7 ± 0.57^†	39.4 ± 0.9^†	35.0 ± 0.8^†
PLT	203.2 ± 5.8^*	707.2 ± 7.8^†	792.5 ± 5.2^†	1100.3 ± 9.2^‡

---

Fig. 1 Effects of individual and combined exposure to arsenic and fluoride on (A) blood ROS and (B) blood GSH level in mice. Abbreviations used and units: ROS—reactive oxygen species as FIU; GSH—glutathione as mg ml⁻¹. Values are mean ± SE; n = 5. *, †, ‡—Differences between values with matching symbol notations within each bar are not statistically significant at 5% level of probability.

Effect on blood G6PD activity

Fig. 2 depicts the G6PD electrophoretic pattern in erythrocytes of normal and exposed animals. Decreased band intensity was observed in lane 2 and lane 3 (i.e. arsenic and fluoride alone exposed animals) indicating a depleted antioxidant system. The inhibition was more pronounced in arsenic exposed animals compared to fluoride exposed animals. Animals co-exposed to arsenic and fluoride showed a decreased band intensity compared to normal however, it was less pronounced compared to individual exposure to arsenic.

Fig. 2 Effect of chronic arsenic and fluoride exposure on G6PD activity in erythrocytes in mice as analyzed by Scion image analyzer software. Lane 1: normal; lane 2: arsenic; lane 3: fluoride; lane 4: arsenic + fluoride. Values are mean ± SE; n = 5; *P < 0.05 compared to normal as evaluated by the student ‘t’ test.

Effect on single strand breaks in circulating lymphocytes

Single strand breaks in circulating lymphocytes of normal and exposed animals were determined through the comet assay and the results are shown in Fig. 3A and B. There were no comets seen in lymphocytes of normal animals suggesting an integrated DNA structure. Chronic arsenic or fluoride exposed cells showed typical comets with long tails indicating heavy DNA damage. Co-administration of arsenic and fluoride showed less pronounced damage to DNA as compared to their individual exposure indicated by comparatively reduced comet tail length.

Fig. 3 (A and B) Effect of chronic arsenic and fluoride exposure on the comet assay for DNA damage in lymphocytes in rats, 40×. (A) normal, (B) arsenic, (C) fluoride, (D) arsenic + fluoride. Note decreased tail length in D compared to B and C.

Effect on liver and brain oxidative stress variables

Exposure to arsenic and fluoride alone significantly increased ROS (Fig. 4A) and TBARS (Fig. 4B) levels in liver and brain. Co-exposure to arsenic and fluoride did not produce reactive oxygen species at higher concentration compared to normal animals but it led to lipid peroxidation in liver.

Fig. 4 Effects of individual and combined exposure to arsenic and fluoride on liver and brain (A) ROS and (B) TBARS level in mice. Abbreviations used and units: ROS—reactive oxygen species as nmoles per min per mg of protein; TBARS—thiobarbituric acid reactive substance as μg g⁻¹ tissue. Values are mean ± SE; n = 5. *, †, ‡—Differences between values with matching symbol notations within each bar are not statistically significant at 5% level of probability.

Glutathione plays a major role in the detoxification of toxicants. We examined reduced and oxidized glutathione levels in terms of the GSH ∶ GSSG ratio in tissues. The GSH ∶ GSSG ratio showed significant depletion in animals exposed to arsenic alone. The GSH ∶ GSSG ratio was more pronounced in the fluoride individual exposed group as compared to arsenic alone. No additive effects of arsenic and fluoride were evident during co-exposure to arsenic and fluoride. Antioxidant enzymes SOD, catalase and GPx are the first line of defence against toxicants. Exposure to arsenic alone significantly decreased liver SOD, catalase and GPx activity however; in brain SOD and GPx activity increased and catalase activity remain unaltered. Exposure to fluoride alone significantly decreased liver SOD and catalase activity however, brain SOD and GPx activity increased significantly. Co-exposure to arsenic and fluoride significantly decreased liver SOD and catalase activity while brain SOD activity increased significantly (Table 2).

Table 2 Effects of individual and combined exposure to arsenic and fluoride on liver and brain biochemical variables in mice

	Normal	As	F	As + F
Abbreviations used and units: GSH—reduced glutathione as mg g^−1 tissue; GSSG—oxidized glutathione as mg g^−1 tissue; SOD—superoxide dismutase as units per min per mg protein; CAT—catalase as unit per min per mg protein; GPx—glutathione peroxidase as μg per min per mg protein. Values are mean ± SE; n = 5. ^*, †, ‡Differences between values with matching symbol notations within each bar are not statistically significant at 5% level of probability.
Liver
GSH ∶ GSSG	14.2 ± 0.56^*	8.6 ± 0.25^†	12.5 ± 0.13^‡	9.5 ± 0. 23^†
SOD	1.41 ± 0. 09^*	0.55 ± 0. 03^†	0.87 ± 0. 04^‡	0.58 ± 0.08^†
Catalase	5.91 ± 0.12^*	4.04 ± 0.14^†	4.84 ± 0. 05^†	4.11 ± 0. 14^†
GPx	0.74 ± 0.14^*	0.58 ± 0.03^†	0.66 ± 0.04^*	0.72 ± 0.04^*
Brain
GSH ∶ GSSG	8.4 ± 0.13^*	7.4 ± 0.15^†	7.9 ± 0.19^*	7.6 ± 0.18^*
SOD	2.13 ± 0.04^*	2.8 ± 0.07^†	6.31 ± 0.06^‡	5.72 ± 0.03^‡
Catalase	4.19 ± 0.14^*	4.51 ± 0.09^*	4.01 ± 0.12^*	3.71 ± 0.04^†
GPx	1.04 ± 0.05^*	1.28 ± 0.04^†	1.27 ± 0.08^†	1.03 ± 0.05^*

---

Effect on liver and brain ATPase activity

ATPase is important for import of metabolites necessary for cell metabolism and export of toxins. In arsenic and fluoride alone exposed animals, ATPase activity increased significantly in liver and brain (Fig. 5). In brain, the increase was more pronounced in fluoride alone and arsenic–fluoride co-exposed animals compared to arsenic alone.

Fig. 5 Effects of individual and combined exposure to arsenic and fluoride on liver and brain ATPase activity in mice. Units: value of ATPase activity expressed as nmol of Pi liberated per min per mg of protein. Values are mean ± SE; n = 5. *, †, ‡—Differences between values with matching symbol notations within each bar are not statistically significant at 5% level of probability.

Effect on arsenic and fluoride concentration

Arsenic concentration in blood, brain and liver of arsenic exposed animals increased significantly as compared to normal animals (Fig. 6A) however, during co-exposure to arsenic and fluoride it was significantly less pronounced as compared to individual exposure to arsenic. Similarly we observed significantly increased concentration of fluoride in blood and soft tissues of fluoride alone exposed animals as compared to normal animals but co-exposure to arsenic and fluoride showed significantly decreased fluoride concentration as compared to their fluoride alone exposure (Fig. 6B).

Fig. 6 Effects of individual and combined exposure to arsenic and fluoride on (A) arsenic and (B) fluoride concentration in blood, liver and brain of mice. Units: arsenic in blood as ng ml⁻¹; arsenic in liver and brain as μg g⁻¹ wet tissue; fluoride in blood as μg ml⁻¹; fluoride in liver and brain as μg g⁻¹ tissue. Values are mean ± SE; n = 5. *, †, ‡—Differences between values with matching symbol notations within each bar are not statistically significant at 5% level of probability.

Histopathological observations

Histopathological changes in liver following exposure to arsenic and fluoride individually or in combination are shown in Fig. 7A–D and severity of lesions is summarized in Table 3. Histology of control mice liver showed normal morphology of hepatic parenchyma, hepatic lobules and hepatocytes arranged in cords radiating from the central canal (Fig. 7A). Liver sections of arsenic alone exposed mice showed vacuolar degeneration (Fig. 7B) and desquamation of endothelium of the central canal. Multiple foci of periportal and midzonal necrosis along with infiltration of polymorphonuclear cells were also observed. Liver of fluoride alone exposed mice showed distortion of normal hepatic morphology, cytoplasmic vacuolation, pyknotic hepatocytes (Fig. 7C), and areas of sporadic necrosis surrounded by the infiltrating polymorphonuclear cells with desquamation of endothelium of the central canal. Vacuolar degeneration in hepatocytes of arsenic or fluoride exposed mice was confirmed as a fatty change (Fig. 8A), depicted by red stained fine fat droplets (Fig. 8B and C) distributed homogenously in hepatocytes observed in cryosections of formalin fixed liver tissues. When animals were co-exposed to arsenic and fluoride, liver histology showed mild degeneration of hepatic parenchyma and very few necrotic foci limited to the periportal region only (Fig. 7D). Occurrence of fat droplets was also reduced (Fig. 8D) in animals having co-exposure to arsenic and fluoride. Animals exposed to arsenic or fluoride alone or in combination showed both oncotic (cellular swelling) and apoptotic types of cell injury; apoptotic cells were confirmed by the TUNEL assay. Prevalence of TUNEL positive cells was higher in arsenic or fluoride exposed mice compared to mice having co-exposure and control mice rarely showed TUNEL positive cells (Fig. 9A–E).

Fig. 7 Photomicrographs of control and arsenic or fluoride alone or in combination exposed mice liver, H & E, 40×. (A) Control mice showing normal hepatic chord, hepatocytes, central canal and Kupffer cells. (B) Section of mice treated with arsenic showing hepatocyte vacuolation (arrows), ballooning of hepatocytes and necrosis (arrow heads). (C) Exposure with fluoride showing hepatocytic vacuolation (arrows), cells showing pyknotic nuclei (arrow heads) with ballooning of hepatocytes and karyolysis. (D) Arsenic and fluoride co-exposure showing pyknotic nuclei (arrow), hepatocellular degeneration and necrotic hepatocytes. Note reduced severity of hepatic lesions in D as compared to B and C.

Table 3 Histopathological grading of liver and brain following individual and combined exposure to arsenic and fluoride in mice

S. No.	Lesions	Control	As	F	As + F
N = 6. −, nil; +, minimal (<12%); ++, mild (<22%); +++, moderate (<45%) and ++++, severe (>45%).
Liver
1	Hepatocellular degeneration	−	++++	++++	++
2	Hypertrophy of Kupffer cells	−	++	++	+
3	Hepatocellular vacuolation	−	+++	++++	+
4	Necrosis	−	+++	+++	++
5	Polymorphonuclear cell infiltration	−	+++	+++	++
6	Pyknosis	−	+++	+++	++
7	Desquamated endothelium of blood vessels	−	+++	+++	+++
Brain
1	Neuronal degeneration	−	+++	+++	++
2	Neuronal vacuolation	−	+++	+++	+
3	Gliosis	−	++	++	+++
4	Pyknosis of neurons	−	+++	+++	+
5	Edema	−	+++	++	+
6	Necrosis	−	+++	+++	++

---

Fig. 8 Oil red O staining of lipid droplets in hepatocytes (arrows) of the liver of mice, 40×. (A) Control without any exposure. (B) Arsenic exposed mice. (C) Fluoride exposed mice. (D) Co-exposure with arsenic and fluoride. Note decreasing severity of fatty change in the C > B > D > A order.

Fig. 9 In situ detection of fragmented DNA [deoxyribonucleotidyl transferase-mediated dUTP-FITC nick-end labeling (TUNEL) assay] in the liver of mice, 40×. (A) Positive control (DNA damage was induced in the liver section by the application of HCl) showing TUNEL positive cells with yellowish green fluorescence. (B) Control (without any exposure) mice liver sections not showing any yellowish green fluorescence. (C) and (D) Arsenic and fluoride exposed mice liver sections showing yellowish green TUNEL positive cells (arrows). (E) Arsenic and fluoride coexposed mice liver sections showing yellowish green TUNEL positive cells (arrow). Note decrease in the intensity of fluorescence in E as compared to C and D.

Histopathological changes in brain (cortex and hippocampus regions) following exposure to arsenic and fluoride individually or in combination are shown in Fig. 10A–H. The cortex region of control mice brain showed normal neurons, glial cells arranged in several layers (Fig. 10A). Mice with arsenic exposure showed moderate gliosis (proliferation of glial cells), shrunken pyknotic neurons along with central chromatolysis, edema and mild to moderate neuronal vacuolation (Fig. 10B). Animals with fluoride exposure showed moderate chromatolysis, gliosis and edema along with necrotic and degenerative changes (Fig. 10C). Animals with arsenic and fluoride co-exposure showed moderate gliosis (Fig. 10D), however other degenerative lesions were less intense compared to arsenic or fluoride exposure alone.

Fig. 10 Photomicrographs of control and treated mice brain (cortex, A–D; hippocampus region, E–H), H & E, 40×. (A) Control mice showing various types of normal neurons, glial cells arranged in several layers. (B) Arsenic-exposed mice brain showing severe chromatolysis (arrows), gliosis and edema in cortex. (C) Mice brain exposed to fluoride showing moderate necrotic (arrow) and degenerative changes. (D) Arsenic and fluoride co-exposure showing marked gliosis (arrows) and reduced severity of necrotic and degenerative lesions in brain cortex compared to B and C. (E) Control mice showing a normal glial cell layer, molecular layer and Purkinje layer. (F) Arsenic-exposed mice brain showing severe chromatolysis of nuclear material (arrow) and most of the Purkinje neurons are necrotic. (G) Section of mice exposed with fluoride showing necrotic degeneration of Purkinje neurons, lysis of glial cells (arrows). (H) Arsenic and fluoride co-exposure showing reduced severity of necrotic (arrow) and degenerative lesions in hippocampus of brain compared to F and G.

The hippocampal region of control mice showed normal cellular composition in all the three layers (Fig. 10E). Arsenic exposed mice brain showed moderate neuronal degeneration with shrinkage and pyknosis in pyramidal neurons (Fig. 10F). Severity of neuronal degeneration and pyknosis in the hippocampal region varied from mild to moderate in fluoride exposed mice brain (Fig. 10G). On co-exposure to arsenic and fluoride minimal lesions were noticed as shown in Table 3.

Spectroscopic observations

The characteristic peaks of FT-IR spectra of control and exposed mice liver are shown in Table 4 and brain are shown in Table 5. A broad intense band at 3396.4 cm⁻¹ in liver and at 3385.7 cm⁻¹ in brain was observed which is assigned to amide A, –NH and –OH stretching vibrations. Animals exposed to arsenic or fluoride alone or in combination showed increase in this peak in both liver and brain. In liver of control animals a sharp peak was obtained at 1455.4 cm⁻¹, which decreased significantly in all exposed groups. A significant peak of amide III –NH bending was observed only in liver of control animals at 1239.9 cm⁻¹ and in brain at 1257.4 cm⁻¹ while it was absent in all exposed animals. A significant peak of the carbonate domain was observed in liver of control animals at 836.0 cm⁻¹ and it increased further in all exposed animals.

Table 4 Characteristics IR peaks (cm⁻¹) of liver following individual and combined exposure to arsenic and fluoride in mice

Groups
Normal	3396.4	1654.4	1455.4	1239.9	836.0	582.7
As	3419.1	1652.7	1410.9	ND	859.9	581.5
F	3418.8	1653.9	1408.2	ND	862.8	583.9
As + F	3410.9	1653.7	1408.1	ND	861.7	584.9

---

Table 5 Characteristics IR peaks (cm⁻¹) of brain following individual and combined exposure to arsenic and fluoride in mice

Groups
Normal	3385.7	1654.1	1417.0	1257.4	865.9	581.8
As	3399.9	1653.6	1419.0	ND	863.5	583.8
F	3399.1	1653.4	1418.8	ND	832.9	584.0
As + F	3388.5	1653.7	1419.1	ND	833.3	585.6

---

Discussion

Arsenic and fluoride have been recognized globally as major environmental pollutants and a threat to human health on chronic exposure, affecting large human population especially in the developing countries. These toxicants are being extensively studied individually for their toxic effects and disease implications on acute and chronic exposure. Human exposure to arsenic or fluoride has mainly been associated with contaminated drinking water.² These well known ground water contaminants were investigated in the present study for their chronic effects when co-ingested through drinking water in mice. The study suggests that despite significant adverse effects of arsenic and fluoride individually after chronic exposure, these toxicants in combination were evident to be either less toxic or their toxicities were comparable to those of individual effects. This can be attributed to a decreased systemic absorption of the arsenic–fluoride metal complex that may be formed in the case of co-exposure. Arsenic was found to be more toxic than fluoride at the same dose and duration. Oxidative stress mediated or direct metal-induced genotoxicity may be a crucial underlying mechanism in toxic manifestations resulting from arsenic and fluoride co-exposure.

Toxicants may interact to produce additive, synergistic or antagonistic response at the biological site. These interactions may be physical or physiological. Results from the present study reveal that during co-exposure, arsenic and fluoride concentrations in blood, liver and brain were less pronounced compared to their individual exposure. The possibility of arsenic–fluoride complex formation has previously been discussed by us.^8,9 The complex thus formed may affect the absorption and distribution of these toxicants in the body. The present study for the first time investigates arsenic and fluoride accumulation following long term co-exposure. The present study further reports more than 50% reduction in arsenic or fluoride accumulation when co-administered, which obviously forms the basis for the lowered toxicity.

Oxidative stress, postulated as a major underlying mechanism of arsenic or fluoride-induced toxicity, was evaluated in the present study to compare effects after co-exposure. Oxidative stress arises when production of ROS in the cells due to endogenous or exogenous (toxicant exposure) factors exceeds their natural antioxidant defense capacity, causing damage to macromolecules such as DNA, proteins and lipids.^47–49 Free radical generation is suggested to be the underlying cause responsible for various toxic manifestations during chronic arsenic or fluoride exposure.^7,47 The antioxidant defense system of the body operates for scavenging ROS to prevent oxidative stress. Intracellular antioxidant enzymes, SOD, catalase, and GPx, are considered to be the first line of cellular defense that protects the biological macromolecules (like DNA, proteins, etc.) from oxidative damage. The second line of cellular defense includes glutathione antioxidant systems that play a crucial role against free radicals and other oxidant species. With its –SH groups, GSH functions as a catalyst in the disulfide exchange reactions. During oxidative stress the –SH group becomes oxidized to form a disulfide link known as GSSG.

Results from the present study demonstrated increase in generation of ROS along with decreased GSH level in blood of arsenic or fluoride exposed animals suggesting a higher degree of oxidative stress and damage in the cells. Increased ROS and TBARS level was also observed in liver and brain of arsenic or fluoride exposed animals accompanied by decreased GSH/GSSH ratio. Ercal et al. (1996) suggested that the GSH/GSSG ratio could be a sensitive indicator of oxidative stress.⁵⁰ It has been suggested that reactive intermediates might react with GSH either by a direct chemical reaction or via glutathione-S-transferase mediated reactions and can be converted into oxidized glutathione (GSSG).⁵¹ When the ratio of oxidized glutathione exceeds the reduction capacity of glutathione reductase, the oxidized glutathione (GSSG) is actively transported out of the cell and is thereby lost. This process might explain the reduction of the GSH/GSSG ratio after arsenic or fluoride exposure. The higher reduction of the GSH ∶ GSSG ratio observed in liver as compared to brain is due to higher arsenic accumulation in liver and depletion of glutathione in hepatic arsenic metabolism. It also indicates liver to be more susceptible to arsenic exposure. To recycle GSSG, cells utilize the NADPH dependent GSH reductase. In return, NADPH is supplied by glucose-6-phosphate dehydrogenase (G6PD).⁵² Glucose-6-phosphate dehydrogenase (G6PD) is the first enzyme of the pentose phosphate pathway. It regulates the operation of the hexose monophosphate shunt (HMS),⁵³ the pathway that produces NADPH and pentose phosphates. It supplies extra mitochondrial NADPH to the cells through the oxidation of glucose-6-phosphate to 6-phosphogluconate. This NADPH maintains the GSH at a constant level. In this study, a significant decrease in blood G6PD activity in arsenic or fluoride exposed animals suggests a decrease in the NADPH level, which reduces the conversion of oxidized glutathione into reduced glutathione. During concomitant exposure to arsenic and fluoride we observed a significantly decreased level of blood GSH, decreased GSH ∶ GSSG ratio in liver and brain accompanied by an increased TBARS level in liver as compared to control animals. However, we did not obtain more pronounced toxicity during concomitant exposure as compared to individual exposure to arsenic and fluoride suggesting no synergism between these toxicants. Both arsenic and fluoride are known to cross the blood brain barrier but during combined exposure an almost normal level of ROS and TBARS in brain suggests some interaction between these toxicants thereby inhibiting their free access in brain tissue. A number of reports suggest formation of an arsenic–fluoride complex either via direct reaction of elemental arsenic with fluorine or with arsenic trioxide.⁵⁴ Concomitant exposure to arsenic and fluoride was also found to decrease blood and soft tissues arsenic or fluoride concentration suggesting some antagonism between them.

Furthermore arsenic or fluoride, individually, altered activities of antioxidant enzymes. SOD and catalase activities in liver of arsenic or fluoride exposed animals decreased significantly, while in brain, activities of SOD and GPx increased and catalase remain unaltered. This decrease or increase in the activities of enzymes can be explained by their consumption and induction during the conversion of free radicals into less harmful metabolites. The decrease in the activities of enzymes suggests their insufficiency in the compensation of free radicals generated on arsenic or fluoride exposure and makes the tissue more susceptible to biochemical injury.^8,55 Increased brain SOD and GPx activities during arsenic–fluoride co-exposure seem reasonable to suggest that increasing intracellular antioxidant levels will have preventive effects on arsenic–fluoride poisoning.^56,57

It is clear that arsenic or fluoride exposure is associated with increase in intracellular ROS. However, the mechanism of ROS generation is not clear. Yih et al. (1991) showed that arsenic treatment destroys membrane structure of mitochondria and decreases the ATP level.⁵⁸ Mishra et al. (2008) reported increased mitochondrial membrane potential due to arsenic exposure in guinea pigs, which confirms mitochondrial membrane disruption.⁵⁹ Fluoride is also known to cause injury to cell membrane and to affect the activity of membrane bound enzymes like ATPase by disturbing membrane fluidity and membrane integrity and altering its permeability.⁶⁰ In the present experiment increased activity of ATPase on arsenic or fluoride exposure alone and in combination in both liver and brain could impart the excessive breakdown of ATP in order to maintain energy requirement of cells. Alternatively, generation of ROS may occur through the inhibition of GPx activity⁶¹ and the binding of arsenite to glutathione.⁵⁹

We also observed that individual exposure to arsenic or fluoride caused significant DNA damage in terms of comet tail in lymphocytes. However, in arsenic and fluoride co-exposed animals the damage was less pronounced compared to their individual exposures indicated by decreased comet tail. It has been reported earlier that the comet assay could be used as an indicator of arsenic or fluoride poisoning.^62,63 However, most of these studies involving humans and animals were reported after acute and sub-acute exposure to these toxicants and the present study confirms these results in chronically exposed animals. We noted a loss of DNA integrity due to single strand breaks. Recent studies have proposed two modes of action for arsenic-induced DNA damage: (i) inhibition of various enzymes involved in DNA repair and expression e.g. polyADP-ribose polymerase-I (PARP-I), an important DNA repair enzyme,^64–66 and (ii) induction of ROS capable of inflicting DNA damage.¹⁴ A possible mechanism of DNA damage induced by fluoride is as follows: (i) fluoride has a dense negative charge and is biochemically very active. Thus it can also have a direct effect on DNA because of its strong affinity for uracil and amide bonds by –NH⋯F– interactions that can induce the rupture of hydrogen bonds in the base pairing of adenine and thymine, resulting in disturbance of the synthesis of DNA and increasing error frequency of linkage between basic groups in the process of DNA replication,²⁰ (ii) fluoride can combine stably with DNA by covalent bonding, affecting the normal structure of DNA, (iii) fluoride can induce the production of free radicals, which can damage DNA strands directly or by lipid peroxidation initiated by free radicals,⁶⁷ (iv) fluoride may depress enzyme activity, such as with DNA polymerase, which might further affect the process of DNA replication or repair and thereby damage DNA.⁶⁸

Increased concentration of ROS diminished effectiveness of the antioxidant defense system and the decreased energy level in cells might be associated with tissue damage which may eventually lead to cell death. Chronic arsenic or fluoride toxicity is known to cause metabolic lesions and cellular deformities in various soft tissues including liver and brain.^69,70 In the present study histopathological examination of liver of arsenic or fluoride exposed animals showed signs of hepatocellular degeneration, vacuolation, inflammation and pyknosis, which are in accordance with previous findings.^2,57 Inhibition of carnitine palmitoyltransferase I, acyl–coenzyme A dehydrogenase, decreased adenosine triphosphate (ATP) levels in hepatocytes and lipid peroxidation may account, in part, for deposition of fat droplets in hepatocytes.⁷¹ Accumulation of fat droplets in hepatocytes of arsenic or fluoride treated rats was rarely noticed on arsenic and fluoride co-exposure. An additional study is required for elaborating the mechanism involved in lipid accumulation after arsenic or fluoride exposure and its sparing effect after their co-exposure in rats. Histology of brain of arsenic or fluoride exposed animals showed severe degenerative and necrotic changes. Varner et al. (1998) also showed vacuolation, and decreased neuronal density in animals exposed to fluoride.⁶⁹ Concomitant exposure to arsenic and fluoride also caused cellular damage however, it was less pronounced as compared to their individual exposure. Histopathological findings also supported the results of biochemical observations along with arsenic or fluoride concentration in soft tissues.

Prelethal cell death reactions have been categorized into oncosis and apoptosis. Oncosis involves swelling of the cells prior to death, however, in apoptosis, the cells shrink, showing cytoplasmic blebs and condensation of chromatin material, ultimately leading to fragmentation.⁷² The deoxyribonucleotidyl transferase (TDT)-mediated dUTP-digoxigenin nick-end labeling (TUNEL) assay is utilized for observing apoptosis. In the present study, we have observed both oncotic and apoptotic types of prelethal cell injury in arsenic and fluoride alone or in co-exposed groups. Pathways for arsenic-induced apoptosis include ROS, mitochondrial disruption, caspase activation, p53, and the MAPK signaling pathway. Binding of arsenic to the SH-group of glutathione (GSH) results in an accumulation of intracellular ROS leading to activation of caspases and ultimately causes apoptosis.⁷³ Fluoride induces apoptosis by elevating oxidative stress-induced lipid peroxidation, thus causing mitochondrial dysfunction and the activation of downstream pathways.^74,75 Results of arsenic- and fluoride-induced apoptosis are in agreement with the findings of Bashir et al. (2006) and Barbier et al. (2010).^57,76 However, reduced prevalence of TUNEL positive cells in co-exposed groups may be attributed to (i) low concentration of arsenic and fluoride at the tissue site and (ii) possible reduction in the active site of the complex for interaction with DNA. Further exploration on the concepts may be recommended.

Further to find out the mechanism of arsenic- or fluoride-induced cellular damage, we recorded FT-IR spectra of liver and brain tissues of normal and exposed animals. FT-IR spectroscopy is the most useful tool for the analysis of secondary structure of proteins.⁷⁷ A significant peak of amide III (–NH) bending was observed at 1239.9 cm⁻¹ in liver and at 1257.4 cm⁻¹ in brain of control animals. This peak was not detected in animals exposed to arsenic or fluoride alone as well as in combination in both liver and brain. It suggests the possibility of hydrogen bonding between fluoride ions and the –NH group.²⁰ Arsenic (III) owing to its electrophilic property might be interacting with the –NH group which leads to denaturation of protein. The band area corresponding to the amide A band was observed at 3396.4 cm⁻¹ in liver and at 3385.7 cm⁻¹ in brain of control animals. It increased on arsenic or fluoride exposure alone and in combination in both liver and brain. In arsenic alone exposed liver the shift was 22.76 cm⁻¹ and in fluoride alone exposed liver it was 22.42 cm⁻¹ while in arsenic–fluoride co-exposed liver the shift was 14.54 cm⁻¹, which was less pronounced as compared to their individual exposure. Similarly in arsenic alone exposed brain the shift was 14.22 cm⁻¹ and in fluoride alone exposed brain it was 13.38 cm⁻¹ while in arsenic–fluoride co-exposed brain the shift was only 2.83 cm⁻¹ suggesting some antagonistic behavior during arsenic–fluoride co-exposure. The large shift due to arsenic or fluoride exposure might imply a variation in the strength of protein and amide hydrogen bonding due to changes in the plasma chemistries. Thus, in continuation of our previous reports suggesting arsenic– or fluoride–DNA binding via the carbonyl or amide binding site, the present study indicates the effect on cellular protein.⁸

Conclusions

In conclusion, the study confirms that arsenic or fluoride accumulates in blood, brain and liver causing oxidative stress and inhibiting auto-oxidation mechanisms, thereby resulting in oxidative damage of DNA and soft tissues. Arsenic or fluoride also inhibits enzymes involved in energy production thus leading to a cascade resulting in altered functions of liver and brain. However, following co-exposure we suggest formation of an inert or less toxic complex that may not easily cross physiological barriers rendering the toxin unavailable for gastric absorption and entering cells. Low arsenic and fluoride concentration and reduced toxicity in blood, liver and brain, following chronic co-exposure, compared to those in individual exposures support the ‘altered-kinetic hypothesis’. The study also indicates greater susceptibility of liver to arsenic or fluoride than brain, liver being the major metabolic site for arsenic, compared to brain that has low toxicant accessibility due to the blood brain barrier; while showing higher toxic potential of arsenic compared to fluoride. On the basis of the present observation it however may be suggested that co-exposure to arsenic and fluoride may not necessarily lead to synergistic or additive effects but may show certain antagonistic interactions in terms of relieving oxidative stress and genotoxicity.

Acknowledgements

Authors thank Dr R. Vijayaraghavan, Director of the establishment, for his support and encouragement.

References

1. S. Lynn, J. R. Gurr, H. T. Lai and K. Y. Jan, Circ. Res., 2000, 86, 514–519.
2. S. Chouhan and S. J. S. Flora, Indian J. Exp. Biol., 2010, 48, 666–678.
3. X. S. Liu, M. Athar, I. Lippal, C. Waldren and T. K. Hei, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 1643–1648.
4. Y. M. Shivarajashankara, A. R. Shivashankara, P. G. Bhat and S. H. Rao, Indian J. Exp. Biol., 2003, 41, 857–860.
5. S. J. S. Flora, V. Pachauri, M. Mittal and D. Kumar, BioMetals, 2011, 24, 615–628.
6. E. Garcia-Chavez, A. Santamaria, F. Diaz-Barriga, P. Mandeville, B. I. Juarez and M. E. Jimenez-Capdeville, Brain Res., 2003, 976, 82–89.
7. S. J. S. Flora, S. Bhadauria, S. C. Pant and R. K. Dhaked, Life Sci., 2005, 77, 2324–2337.
8. S. J. S. Flora, M. Mittal and D. Mishra, J. Neurol. Sci., 2009, 285, 198–205.
9. M. Mittal and S. J. S. Flora, Chem.-Biol. Interact., 2006, 162, 128–139.
10. M. Mittal and S. J. S. Flora, Drug Chem. Toxicol., 2007, 30, 263–281.
11. S. Shila, V. Kokilavani, M. Subathra and C. Panneerselvam, Toxicology, 2005, 210, 25–36.
12. K. P. Singh, R. Kumari, J. Treas and J. W. Dumond, Chem. Res. Toxicol., 2011, 24, 340–349.
13. A. Basu, J. Mahata, S. Gupta and A. K. Giri, Mutat. Res., 2001, 488, 171–194.
14. T. Schwerdtle, I. Walter, I. Mackiw and A. Hartwig, Carcinogenesis, 2003, 24, 967–974.
15. D. Li, K. Morimoto, T. Takeshita and Y. Lu, Environ. Health Prev. Med., 2001, 6, 27–32.
16. M. Wei, H. Wanibuchi, K. Morimura, S. Iwai, K. Yoshida, G. Endo, D. Nakae and S. Fukushima, Carcinogenesis, 2002, 23, 1387–1397.
17. Z. Liu, J. Shen, J. M. Carbrey, R. Mukhopadhyay, P. Agre and B. P. Rosen, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 6053–6058.
18. C. Yedjou, P. Tchounwou, J. Jenkins and R. McMurray, J. Hematol. Oncol., 2010, 26, 3–28.
19. Z. X. Shen, G. Q. Chen, J. H. Ni, X. S. Li, S. M. Xiong, Q. Y. Qiu, J. Zhu, W. Tang, G. L. Sun, K. Q. Yang, Y. Chen, L. Zhou, Z. W. Fang, Y. T. Wang, J. Ma, P. Zhang, T. D. Zhang, S. J. Chen, Z. Chen and Z. Y. Wang, Blood, 1997, 89, 3354–3360.
20. Y. M. Li, N. A. Heerema, A. J. Dunipace and G. K. Stookey, Mutat. Res., 1987, 192, 191–201.
21. D. Shanthakumari, S. Srinivasalu and S. Subramanian, Toxicology, 2004, 204, 219–228.
22. D. Slamenova, A. Gabelova and K. Ruppova, Mutat. Res., 1992, 279, 109–115.
23. N. H. Kleinsasser, H. Weissacher, B. C. Wallner, E. R. Kastenbauer and U. A. Harreus, Laryngol., Rhinol., Otol., 2001, 80, 187–190.
24. D. A. Ribeiro, M. E. Marques, G. F. Assis, A. Anzai, M. L. Poleti and D. M. Salvadori, Caries Res., 2004, 38, 576–779.
25. N. J. Chinoy and T. N. Patel, Fluoride, 2001, 34, 9–20.
26. Z. Guan, K. Xiao, X. Zeng, Y. Long, Y. Cheng and S. Jiang, Arch. Toxicol., 2000, 74, 602–608.
27. N. J. Chinoy and M. R. Memon, Fluoride, 2001, 34, 21–33.
28. N. J. Chinoy and T. N. Patel, Fluoride, 1999, 32, 215–229.
29. J. Liu, B. S. Zheng and A. Vasken, Environ. Health Perspect., 2002, 110, 119–122.
30. A. Shashi and S. P. Thapar, Fluoride, 2001, 34, 43–50.
31. S. D. Shah and N. J. Chinoy, Fluoride, 2004, 37, 162–171.
32. F. Kassie, W. Parzefall and S. Knasmuller, Mutat. Res., 2000, 463, 13–31.
33. G. Speit, S. Hanelt, R. Helbig, A. Seidel and A. Hartmann, Toxicol. Lett., 1996, 88, 91–98.
34. G. L. Ellman, Arch. Biochem., 1959, 82, 70–77.
35. D. J. Jollow, J. R. Mitchell, Z. Zamppaglione and J. R. Gillette, Pharmacology, 1974, 11, 151.
36. D. J. Socci, K. B. Bjugstad, H. C. Jones, J. V. Pattisapu and G. W. Arendash, Exp. Neurol., 1999, 155, 109–117.
37. L. Ornstein, Ann. N. Y. Acad. Sci., 1964, 121, 321–323.
38. A. Scholl and S. Anders, Arch. Genet., 1973, 46, 121–129.
39. N. P. Singh, M. T. McCoy, R. R. Tice and E. L. Schneider, Exp. Cell Res., 1988, 171, 184–191.
40. P. V. L. Rao, R. Vijayaraghavan and A. S. B. Bhaskar, Toxicology, 1999, 139, 39–51.
41. P. J. Hissin and R. Hilf, Anal. Biochem., 1976, 74, 214–216.
42. H. Ohkawa, N. Onishi and K. Yagi, Anal. Biochem., 1979, 95, 351–358.
43. P. Kakkar, B. Das and P. N. Viswanathan, Indian J. Biochem. Biophys., 1984, 21, 130–132.
44. H. Aebi, Methods in Enzymology, Academic Press, New York, 1984, pp. 125–126.
45. L. Flohe and W. A. Gunzler, Methods Enzymol., 1984, 105, 114–121.
46. D. E. Griffiths and R. L. Houghton, Eur. J. Biochem., 1974, 46, 157–167.
47. Y. M. Shivarajashankara, A. R. Shivashankara, B. P. Gopalakrishna and R. S. Hanumanth, Fluoride, 2001, 34, 108–113.
48. H. Bartsch and J. Nair, Toxicology, 2000, 153, 105–114.
49. B. Halliwell, Biochem. Soc. Trans., 2007, 35, 1147–1150.
50. N. Ercal, P. Treeratphan, T. C. Hammond, R. H. Mattews, N. Grannemann and D. Spitz, Free Radical Biol. Med., 1996, 21, 157–161.
51. R. A. Zakharyan, E. Wildfang and H. V. Aposhian, Toxicol. Appl. Pharmacol., 1996, 140, 77–84.
52. A. J. Grimes, Human Red Cell Metabolism, Blackwell Scientific, Oxford, 1980, p. 257.
53. L. Luzzatto and G. Battestuzzi, Adv. Hum. Genet., 1984, 14, 217–329.
54. D. R. Aris, C. Knapp, J. Passmore and X. Wang, J. Fluorine Chem., 2005, 126, 1368–1372.
55. P. D. Patel and N. J. Chinoy, Fluoride, 1998, 31, S27–S32.
56. S. Maiti and A. K. Chatterjee, Arch. Toxicol., 2001, 75, 531–537.
57. S. Bashir, Y. Sharma, M. Irshad, S. D. Gupta and T. D. Dogra, Basic Clin. Pharmacol. Toxicol., 2006, 98, 38–43.
58. L. H. Yih, H. M. Huang, K. Y. Jan and T. C. Lee, Cell Biol. Int. Rep., 1991, 15, 253–264.
59. D. Mishra, A. Mehta and S. J. S. Flora, Chem. Res. Toxicol., 2008, 21, 400–407.
60. N. J. Chinoy, A. K. Sharma, T. N. Patel, R. Memon and D. D. Jhala, Fluoride, 2004, 37, 257–263.
61. F. Z. Sheabar and S. Yannai, Vet. Hum. Toxicol., 1989, 31, 528–531.
62. A. S. Andrew, J. L. Burgess, M. M. Meza, E. Demidenko, M. G. Waugh, J. W. Hamilton and M. R. Karagas, Environ. Health Perspect., 2006, 114, 1193–1198.
63. Y. Ge, H. Ning, S. Wang and J. Wanga, Fluoride, 2005, 38, 209–214.
64. M. Klaude, S. Eriksson, J. Nygren and G. Ahnstrom, Mutat. Res., 1996, 363, 89–96.
65. P. J. McCarthy, S. F. Sweetman, P. G. McKenna and V. J. McKelvey-Martin, Mutagenesis, 1997, 12, 209–214.
66. J. W. Yager and J. K. Wiencke, Mutat. Res., 1997, 386, 45–51.
67. A. G. Wang, T. Xia, Q. L. Chu, M. Zhang, F. Liu, X. M. Chen and K. D. Yang, Biomed. Environ. Sci., 2004, 17, 217–222.
68. N. Hayashi and T. Tsutsui, Mutat. Res., 1993, 290, 293–302.
69. J. A. Varner, K. F. Jensen, W. Horvath and R. L. Isaacson, Brain Res., 1998, 16, 284–298.
70. A. R. Shivashankara, Y. M. Shivarajashankara, P. G. Bhat and S. H. Rao, Bull. Environ. Contam. Toxicol., 2002, 68, 612–616.
71. D. Pessayre, A. Berson, B. Fromenty and A. Mansouri, Semin. Liver Dis., 2001, 21, 057–070.
72. T. E. Benjamin, I. K. Berezesky, S. H. Chang and P. C. Phelps, Toxicol. Pathol., 1997, 25, 82.
73. Y. C. Chen, S. Y. Lin-Shiau and J. K. Lin, J. Cell. Physiol., 1998, 177, 324–333.
74. H. Karube, G. Nishitai, K. Inageda, H. Kurosu and M. Matsuoka, J. Dent. Res., 2009, 88, 461–465.
75. C. D. Anuradha, S. Kanno and S. Hirano, Free Radical Biol. Med., 2001, 31, 367–373.
76. O. Barbier, L. Arreola-Mendoza, L. Mar and D. Razo, Chem.-Biol. Interact., 2010, 188, 319–333.
77. A. Dogan, K. Ergen, F. Budak and F. Severcan, Appl. Spectrosc., 2007, 61, 199–203.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c1mt00118c

---