Sodium fluoride induces apoptosis in mouse splenocytes by activating ROS-dependent NF-κB signaling

In this study, we investigated the roles of reactive oxygen species (ROS) and nuclear factor-κB (NF-κB) signaling in sodium fluoride-induced DNA damage and apoptosis in mouse splenocytes. Intragastric administration of 12, 24 or 48 mg/kg sodium fluoride resulted in a time- and dose-dependent increase in DNA fragmentation and apoptosis in mouse splenocytes on days 21 and 42. High ROS levels correlated with increased levels of phosphorylated IκB kinase and NF-κB p65 and decreased levels of inhibitory kappa B protein in splenocytes from mice treated with sodium fluoride. Moreover, splenocytes from sodium fluoride-treated mice showed high expression of pro-apoptotic proteins, including Bim, Bax, Bak, caspase-3 and poly ADP-ribose polymerase, and low expression of the anti-apoptotic proteins BcL-2 and BcL-xL. These results show that sodium fluoride induces apoptosis in mouse splenocytes by enhancing ROS-dependent NF-κB signaling.


INTRODUCTION
Prolonged intake of fluoridated drinking water and food results in dental, skeletal and non-skeletal fluorosis [1,2]. Non-skeletal fluorosis is characterized by severe gastrointestinal, neurological, muscular and hematological symptoms in both humans and animals [3][4][5][6][7]. High dietary fluoride affects humoral and cellular immunity by decreasing levels of serum cytokines and immunoglobulins [8,9]. Moreover, fluoride reduces lymphocytes in the white and red pulp of the spleen [10].
Reactive oxygen species (ROS) are byproducts of cellular metabolism that modulate signaling pathways in response to changes in the intracellular and extracellular environments [11]. High ROS levels induce changes in cellular structure and function as a result of somatic DNA mutations that eventually promote neoplastic transformation [12,13]. Sodium fluoride alters transcription of xenobiotic metabolizing enzymes in zebra fish by increasing ROS [14]. Moreover, fluoride increase the level of ROS contributing to cytotoxicity in human neuroblastoma SH-SY5Y cells [15].
Genotoxic agents induce extensive DNA damage that induces cellular apoptosis [16,17]. Apoptosis is a physiological process that is necessary for organ development, tissue homeostasis, and elimination of defective cells [18]. DNA damage due to fluoride toxicity results in apoptosis of murine lymphocytes [19], human embryo-derived hepatocytes [20], and rat oral mucosal cells and hepatocytes [21,22].
Nuclear factor κB (NF-κB) signaling plays an important role in DNA damage response and apoptosis in diverse cell type [23][24][25][26][27]. Moreover, fluoride promotes expression of NF-κB pathway genes and activation of the NF-κB signaling pathway [28][29][30]. However, the effects of fluoride toxicity in splenocytes are unknown. Therefore, in this study, we investigated the roles of ROS and NF-κB signaling pathway in NaF-induced DNA damage and apoptosis in the mouse splenocytes.

NaF-treated mice show dose-and timedependent increase in DNA fragmentation and apoptosis in splenocytes
Nucleosomal DNA fragmentation by endogenous endonuleases cleaves into 180-200 bp fragments is a characteristic feature of cellular apoptosis [31][32][33]. We investigated status of DNA fragmentation in the splenocytes of mice treated with NaF by agarose gel electrophoresis. We observed dose-and time-dependent increase in DNA fragmentation in splenocytes of mice treated with 24 and 48 mg/kg NaF on day 21; highest DNA fragmentation was observed in the 48 mg/kg group on day 42 ( Figure 1).
TUNEL assays showed increased number of apoptotic cells in the spleens of the 24 and 48 mg/kg NaF-treated mice on day 21 (p < 0.01; Figure 2) and in the 12-, 24-and 48-mg/kg NaF-treated mice on day 42 (p < 0.01; Figure 3) when compared to the controls. This suggested that NaF induced dose-and time-dependent DNA fragmentation and apoptosis in murine splenocytes.

NaF treatment increases ROS levels in mouse splenocytes
Next, we determined intracellular ROS levels in control and NaF-treated murine splenocytes by ROSsensitive fluorescent dye, DCFH-DA. We observed higher ROS levels in the splenocytes from 24 and 48 mg/ kg NaF-treated mice on day 21 and in the 12, 24 and 48 mg/kg NaF-treated mice on day 42 (p <0.01 or p < 0.05; Figure 4).
Effects of NaF on p-IKK, IκB, NF-κB protein and mRNA levels in the mouse splenocytes Next, we analyzed the status of NF-κB signaling pathway in the control and NaF-treated murine spleens. We observed higher phospho-IKK levels in the 48 mg/ kg group on day 21 and in the 24 and 48 mg/kg groups on day 42 than in the control group (p < 0.01 or p < 0.05; Figure 5C). Moreover, IKK mRNA levels were higher in the 24 and 48 mg/kg groups on day 21 and in the 12, 24 and 48 mg/kg NaF-treated groups on day 42 than in the control group (p < 0.01 or p < 0.05; Figure 5D).
The IκB protein and mRNA expression levels were lower (p < 0.01) in the 12, 24 and 48 mg/kg NaF-treated groups than in the control group on days 21 and 42 ( Figure  5E, 5F). NF-κB protein expression levels were higher in the 24 and 48 mg/kg groups than those in the control group on day 21 (p < 0.01; Figure 5G). Moreover, NF-κB mRNA levels were higher in the 48 mg/kg NaF-treated group than in the control on day 21 (p < 0.01; Figure 5H). NF-κB mRNA and protein levels markedly increased in the 12, 24 and 48 mg/kg groups when compared to the control group on day 42 ( p < 0.01 or p < 0.05; Figure 5G, 5H).

Effects of NaF on Bax and Bcl-2 protein and mRNA expression levels in the mouse splenocytes
We observed higher Bax protein and mRNA levels in the 24 and 48 mg/kg groups on day 21 and in the 12, 24 and 48 mg/kg NaF groups on day 42 than in the control group (p < 0.01 or p < 0.05; Figure 6C, 6D). Conversely, Bcl-2 protein levels decreased in the 48 mg/ kg NaF-treated group on day 21 and in the 24 and 48 mg/ kg NaF-treated groups on day 42 when compared to the Oncotarget 114430 www.impactjournals.com/oncotarget controls (p < 0.01; Figure 6E). Moreover, Bcl-2 mRNA levels were lower in the 48 mg/kg NaF-treated group on day 21 and in the 12, 24 and 48 mg/kg NaF-treated groups on day 42 than in the controls ((p < 0.01; Figure 6F). Therefore, the ratio of Bax/Bcl-2 proteins was higher in the 24 and 48 mg/kg NaF-treated groups on day 21 and day 42 than in the control group (p < 0.01; Figure 6G). Meanwhile, the ratio of Bax/Bcl-2 mRNAs was higher in the 48 mg/kg group on day 21 and in the 24 and 48 mg/ kg NaF-treated group on day 42 than in the control group (p < 0.01; Figure 6H).

Effects of NaF on Bak and Bcl-xL protein and mRNA expression levels in the spleen
The Bak protein and mRNA levels were higher in the 48 mg/kg NaF-treated group on day 21 and in the 24 and 48 mg/kg NaF-treated groups on day 42 than in the control group (p < 0.01; Figure 7C, 7D). Conversely, BcL-xL protein expression levels were lower in the 48 mg/kg group on day 21 and in the 24 and 48 mg/kg NaF-treated group on day 42 than in the controls (p < 0.01 or p < 0.05; Figure  7E). Moreover, the Bcl-xL mRNA levels decreased in the Oncotarget 114431 www.impactjournals.com/oncotarget 24 and 48 mg/kg NaF-treated groups on days 21 and 42 when compared to the control group (p < 0.01 or p < 0.05; Figure 7F). Therefore, the ratio of Bak/Bcl-xL proteins was higher in the 48 mg/kg NaF-treated groups on day 21 and in the 24 and 48 mg/kg NaF-treated groups on day 42 than in the controls (p < 0.01; Figure 7G). Furthermore, the ratio of Bak/Bcl-xL mRNAs was higher in the 24 and 48 mg/kg NaF-treated groups than in the control group on days 21 and 42 (p < 0.01 or p < 0.05; Figure 7H).

Effects of NaF on Bim, caspase-3 and PARP protein and mRNA levels in the murine splenocytes
Bim protein levels were higher in the 12, 24 and 48 mg/kg NaF-treated groups, whereas Bim mRNA levels were higher in 48 mg/kg group than in the control group on day 21 (p < 0.01 or p < 0.05; Figure 8C, 8D). Furthermore, Bim protein and mRNA levels increased in the 12, 24 and 48 mg/kg NaF-treated groups when compared to the control group on day 42 (p < 0.01; Figure  8C and 8D). Caspase-3 protein and mRNA expression was higher in the 24 and 48 mg/kg NaF-treated groups on day 21 and in the 12, 24 and 48 mg/kg NaF-treated groups on day 42 than in the control group (p < 0.01 or p < 0.05; Figure 8E, 8F). PARP protein levels increased in the 48 mg/kg group, whereas PARP mRNA levels increased in the 12, 24 and 48 mg/kg groups in comparison to the controls on day 21 (p < 0.01; Figure 8G, 8H). PARP protein and mRNA levels were higher in the 12, 24 and 48 mg/kg NaF-treated groups than in the control group (p < 0.01; Figure 8G, 8H).

DISCUSSION
Fluorosis is a disease caused by biogeochemical factors such as drinking water with high fluoride levels [34]. In China, fluorosis is endemic [35]. Fluoride-related cytotoxicity causes DNA damage and cellular apoptosis in both humans and animals [36][37][38][39][40]. However, the Oncotarget 114432 www.impactjournals.com/oncotarget roles of ROS and NF-κB signaling pathway in fluorideinduced DNA damage and apoptosis are not well studied. Therefore, we investigated the effects of NaF on the status of ROS and NF-κB signaling pathway in the murine splenocytes. We observed a time-and dosedependent increase in ROS and DNA fragmentation in the splenocytes, which correlated with increased apoptosis and induction of NF-κB pathway and pro-apoptotic proteins. Therefore, our study demonstrates that NaF induces DNA damage and apoptosis in murine splenocytes via ROSdependent NF-κB signaling pathway.
ROS are chemically reactive molecular species such as the superoxide anion (O2-), hydrogen peroxide (H2O2) and the hydroxyl radical (HO•) that are generated by incomplete reduction of oxygen [41,42]. They are highly reactive and toxic to cells at high levels, but, at lower levels, they serve as intra-and inter-cellular signaling molecules that modulate cellular function [43]. The role of ROS homeostasis and signaling has been well documented in human ageing and other complex diseases such as cancer [43]. Moreover, ROS play an important role in the activation of NF-κB signaling pathway [44].
NF-κB is a transcription factor that is expressed in all cell types, which regulates many target genes by binding to its cognate DNA elements [44,45]. It remains inactive in the cytoplasm as a complex with IκB [46]. The most common step of activating Rel/NF-kB proteins appears to occur via induced phosphorylation of IκB by a large IκB kinase complex (IKK), and phosphorylation of IκB leads to its degradation by the proteasome [47]. This releases NF-κB from IκB and eventually translocates into the nucleus as a homo-or hetero-dimer to activate transcription of its target genes [44]. Our findings show that fluoride increases ROS and activates the NF-κB signaling pathway.     Oncotarget 114437 www.impactjournals.com/oncotarget DNA repair involves mechanisms that recognize and correct different kinds of DNA damage due to environmental insults [48]. However, extensive DNA damage results in activation of the cell death mechanisms such as apoptosis and necrosis [48]. Previous studies show that ROS and NF-κB signaling mechanisms are involved in DNA damage [27,49,50]. Our study shows that fluoride induces nuclear DNA fragmentation in a timeand dose-dependent manner. This correlates with high intracellular ROS levels and activation of NF-κB signaling pathway. This indicates that the ROS-dependent NF-κB signaling pathway plays an important role in NaF-induced DNA damage in the mouse splenocytes.
Apoptosis is a multi-step process of programmed cell death that is induced by many chemical genotoxins as a result of DNA replications block [51]. The induction and execution of apoptosis requires co-ordinated regulation of the Bcl-2 family proteins and the caspase signaling cascade [52]. The Bcl-2 family includes pro-or antiapoptotic proteins; activation of pro-apoptotic proteins such as Bax, Bak and Bim induces cellular apoptosis [53,54]; induction of anti-apoptotic proteins such as Bcl-2 and Bcl-xL promotes cell survival [55,56]. Activation of effector caspases such as caspase-3 results in cleavage of various cytoplasmic and nuclear proteins like PARP [52]. NF-κB signaling pathway plays a critical role in the regulation of apoptosis [57]. The pro-or anti-apoptotic role of NF-κB depends on the cell type and apoptotic signal [23]. We demonstrate that fluoride treatment increases expression of pro-apoptotic proteins such as Bax, Bak, Bim, caspase-3 and PARP and decreases expression of anti-apoptosis proteins such as Bcl-2 and Bcl-xL The spleen is the largest secondary lymphoid organ and contains one-fourth of the total lymphocytes and it plays an important role in maintaining immune homeostasis [58,59]. Our study shows that NaF induces DNA damage and apoptosis in the splenocytes by upregulating pro-apoptotic Bcl-2 family proteins via ROSinduced NF-κB pathway.

Experimental mice and NaF-treatment groups
Two hundred and forty healthy ICR mice (Experimental Animal Corporation of DOSSY at Chengdu, China) were divided into 4 groups (N = 60) and intragastrically administered 1 ml of 12, 24 or 48 mg NaF (S6776; Sigma Aldrich, UK) or distilled water per 100 g animal weight for 42 days. Spleens were isolated from control and NaF-treated mice after euthanasia on days 21 and 42. Each experiment was repeated eight times. The mice experimental protocols were approved by the Animal Care and Use Committee of the Sichuan Agricultural University.

Detection of DNA damage
80 μl DNA was extracted from 10 mg spleen tissue of each mice by DNA Ladder Extraction Kit with Spin Column (C0008; Beyotime Biotechnology, Shanghai, China) and electrophoresis on 1% agarose gels at 100V for 1 h, stained with ethidium bromide and visualized by gel imaging analysis system.

Detection of apoptosis by TUNEL
Spleen harvested from mice were fixed in 4% paraformaldehyde overnight followed by dehydration in ethanol, embedded in paraffin wax and serially sectioned into 4 μm thick slices. TUNEL assay was carried out with the Situ Cell Death Detection Kit (Cat. No. 11684817980, Roche, Mannheim, Germany) according to manufacturer's protocol. Briefly, tissue sections were rehydrated in a series of xylene and ethanol solutions and digested with 50 μL 10 μg/ml proteinase K in 10 mM Tris/ HCl, pH 7.4-8 for 15 min. Then, the sections were incubated with 3% H2O2 in methanol for 15 min at room temperature to inactivate endogenous peroxidase. The sections were then incubated with a mixture containing biotin-dUTP and terminal deoxynucleotidyl transferase in a humidified chamber for 1 h at 37°C. Tissue specimens were then incubated in Converter-POD (HRP) for 30 min at 37°C and developed with the Diaminobdenzidine (DAB) kit (AR1022; Boster, Wuhan, China). The specimens were counterstained with hematoxylin, dehydrated in ethanol, clarified in xylene, mounted on slides and observed under an Olympus light microscope (Olympus, Shimadzu, Japan). The apoptotic cells containing DNA strand breaks stained brown. The intensity of staining in specimens was determined by using a computer-supported imaging system connected to a light microscope with an objective magnification of 1000×. Apoptotic cells were quantified by the Image-Pro Plus 5.1 software (Madia Cybernetics, Bethesda, MD, USA). Five sections were analyzed for each mice and five fields were measured in each section and averaged.

Analysis of ROS in mouse splenocytes isolated from NaF treated mice
Cellular ROS levels were evaluated with the ROS detection Kit (S0033; Beyotime Biotechnology, Shanghai, China) according to the manufacturer's protocol. Spleens harvested from mice on days 21 and 42 were gently ground and the single cell suspensions were obtained by filtering through a 300-mesh nylon screen. The splenocytes were washed twice with cold 1X phosphate buffered saline (PBS; pH 7. 2-7.4) and resuspended at a concentration of 1 × 10 6 cells/mL in 1X PBS. Then 1 × 10 5 splenocytes in100 μL PBS were stained with 10 μM DCFH-DA for 20 min at 37°C in the dark. Then, the cell suspension was www.impactjournals.com/oncotarget diluted with 400 μL PBS and analyzed in a FACS Calibur (Becton Dickinson, NJ, USA). Splenocytes treated with Rosup agent, which induces ROS was used as a positive control for DCFH-DA staining.

Quantitative RT-PCR of NF-κB pathway and apoptosis pathway mediators
Frozen spleen samples were homogenized with liquid nitrogen with a mortar and pestle and Total RNA was extracted with the RNAiso Plus (9108/9109, Takara, Japan) according to manufacturer's instructions. Then, cDNA was synthesized with the Prim-Script™ RT reagent Kit (RR047A, Takara, Japan) according to the manufacturer's instructions and used as a template for QRT-PCR. QRT-PCR primers were designed with the Primer 5 software and synthesized by Takara Inc. Dalian, China. Primers were designed using Primer 5 and synthesized at Takara (Dalian, China) ( Table 1). The total reaction mixture for real-time PCR was 25 μL, which included 2 μL cDNA, 12.5 μL of SYBR ® Premix Ex TaqTM II (DRR820A, Takara, Japan), 1μL of each primer (10 μM) and 8.5 μL of RNAase-free water. Real-time PCR cycling conditions were 95°C for 3 min followed by 44 cycles of 95°C for 10 s, Tm of the specific primer pair for 30 s, and 72°C for 10 s. Real time PCR was carried out in a C1000 Thermal Cycler (BIO RAD, CA, USA). The melting curve analysis was performed to ensure a single peak for each PCR product. Further, purity of specific PCR products was verified by agarose gel electrophoresis. Mouse β-actin was used as an internal reference. Gene expression at days 21 and 42 were calibrated against the corresponding controls. Relative expression was analyzed by the 2 -ΔΔCT method [60].