MDM2-Mediated p21 Proteasomal Degradation Promotes Fluoride Toxicity in Ameloblasts

Abstract

Fluoride overexposure is an environmental health hazard and can cause enamel and skeletal fluorosis. Previously we demonstrated that fluoride increased acetylated-p53 and its downstream target p21 in ameloblast-derived LS8 cells. However, p21 function in fluoride toxicity is not well characterized. This study seeks to gain a better understanding of how p53 down-stream mediators, p21 and MDM2, respond to fluoride toxicity. LS8 cells were treated with NaF with/without MG-132 (proteasome inhibitor) or Nutlin-3a (MDM2 antagonist). NaF treatment for 2-6 h increased phospho-p21, which can inhibit apoptosis. However, phospho-p21 and p21 were decreased by NaF at 24 h, even though p21 mRNA was significantly increased at this time point. MG-132 reversed the fluoride-mediated p21 decrease, indicating that fluoride facilitates p21 proteasomal degradation. MG-132 suppressed fluoride-induced caspase-3 cleavage, suggesting that the proteasome plays a pro-apoptotic role in fluoride toxicity. NaF increased phospho-MDM2 in vitro and in mouse ameloblasts in vivo. Nutlin-3a suppressed NaF-mediated MDM2-p21 binding to reverse p21 degradation which increased phospho-p21. This suppressed apoptosis after 24 h NaF treatment. These results suggest that MDM2-mediated p21 proteasomal degradation with subsequent phospho-p21 attenuation contributes to fluoride-induced apoptosis. Inhibition of MDM2-mediated p21 degradation may be a potential therapeutic target to mitigate fluoride toxicity.

Keywords: MDM2; ameloblast; fluoride; fluorosis; p21; p53.

Figure 1

Fluoride upregulates p21 mRNA and p21/p-p21 protein in LS8 cells. (a) LS 8 cells were treated with NaF at the indicated concentrations for 24 h and p21 mRNA was then quantified by quantitative polymerase chain reaction (qPCR). Fluoride (5 mM) treatment significantly increased p21 mRNA levels. Gapdh was the internal reference control gene. Data are presented as means ± standard deviation (SD) (** p < 0.01 vs. 0 mM). (b) Cells were treated with NaF (5 mM) for 6 h and phospo-p21 (p-p21; green), nucleus (4′,6-diamidino-2-phenylindole (DAPI); blue) and β-actin (red) were detected by immunocytochemistry. Fluoride increased p-p21 protein levels in LS8 cells. (c) LS8 cells were treated with NaF (5 mM) for the indicated times and p21 (18 kDa) and p-p21 (21 kDa) were detected by Western blots. Fluoride increased p21 and p-p21 protein expression in the early phase (1 to 6 h), which then decreased at 24 h. The numbers show the relative expression normalized by the loading control β-actin (44 kDa). Statistical analysis of relative protein expression of p21 and p-p21 are shown in supplementary Figure S4.

Figure 2

Fluoride increased Ub-p21 binding and MG-132 reversed the fluoride-mediated p21 protein decrease. (a) LS8 cells were treated with NaF (5 mM) for 6 h. Protein was immunoprecipitated using anti-p21 antibody and ubiquitinated-p21 (Ub-p21) was detected in the precipitated fraction by the anti-Ubiquitin antibody. Fluoride treatment increased Ub-p21 levels. IgG was used as the negative control. The numbers show relative protein expression vs. Controls (0 mM NaF). IP lanes were quantified separately from input lanes. (b) LS8 cells were treated with MG-132 (0.5–1.0 μM) for 2 h prior to NaF (5 mM) treatment for 24 h. p21 (18 kDa) and p-p21 (21 kDa) were detected by Western blot. MG-132 reversed the fluoride-induced p21 suppression at 24 h by increasing p-p21 protein levels. The numbers show relative expression normalized by the loading control β-actin (44 kDa). Statistical analysis of relative protein expression of p21 and p-p21 are shown in supplementary Figure S5. (c) Cells were treated with NaF (5 mM) with/without MG132 (0.5 µM) for 24 h and p-p21 (green), nucleus (DAPI; blue) and β-actin (red) were detected by immunocytochemistry. MG-132 treatment increased p-p21 protein levels.

Figure 3

MG-132 attenuated fluoride-induced apoptosis in LS8 cells. LS8 cells were treated with MG-132 (0.5–1.0 μM) for 2 h prior to NaF (5 mM) treatment for 24 h. (a) The Bcl-2/Bax mRNA ratio was quantified by qPCR. NaF significantly decreased the Bcl-2/Bax mRNA ratio (** p < 0.01). (b) MG-132 (1 μM) significantly increased the Bcl-2/Bax ratio compared to NaF treatment alone (* p < 0.05). Data are presented as means ± SD. (c) γH2AX (15 kDa) and cleaved-caspase-3 (17 kDa) were detected by Western blots. MG-132 inhibited fluoride-induced γH2AX protein expression and inhibited caspase-3 cleavage. The numbers show relative expression normalized by the loading control β-actin (44 kDa). Statistical analysis of relative protein expression of cleaved-caspase-3 and γH2AX are shown in supplementary Figure S6.

Figure 4

Fluoride induced Mdm2 mRNA expression and induced p-MDM2 protein levels in LS8 cells. (a) LS8 cells were treated with the indicated concentrations of NaF for 24 h and Mdm2 mRNA was quantified by qPCR. Fluoride (3 mM or 5 mM) significantly increased Mdm2 expression. Gapdh was the internal reference control gene. Data are presented as the mean ± SD (** p < 0.01 vs. 0 mM). (b) Cells were treated with NaF (5 mM) for the indicated times. Whole cell lysates were subjected to Western blot analysis for phospho-MDM2 (p-MDM2 [Ser166]) (90 kDa) and total MDM2 (MDM2) (90 kDa) expression. β-actin (44 kDa) was used as a loading control. The numbers show relative protein expression normalized by the β-actin loading control. Statistical analysis of relative protein expression of MDM2 and p-MDM2 are shown in supplementary Figure S7.

Figure 5

Fluoride increased p-MDM2 protein expression in mouse ameloblasts. Mice were treated with 0 or 150 ppm fluoride in drinking water for 6 weeks. (a) p-MDM2 [Ser185] was detected by immunohistochemistry in maxillary incisors from mice treated with 0 ppm (upper panel) or 150 ppm (lower panel) fluoride. (b) High magnification of p-MDM2 staining in secretory stage (SEC) and maturation stage (MAT) mouse enamel organs. More p-MDM2 was formed in mouse ameloblasts treated with 150 ppm fluoride compared to control ameloblasts (0 ppm). Shown are representative images from three mice. Scale bar represents 20 μm. Brackets denote ameloblasts (Am).

Figure 6

Fluoride induced p53-MDM2 binding and p53 ubiquitination in LS8 cells. LS8 cells were treated with fluoride for 6 h and protein was immunoprecipitated using anti-p53 antibody. p53, MDM2 and ubiquitin were detected by Western blot. (a) NaF increased MDM2-p53 binding and increased amounts of ubiquitinated-p53 (Ub-p53). (b) MDM2 antagonist Nutlin-3a suppressed fluoride-induced MDM2-p53 binding and decreased Ub-p53 levels. Control IgG was used as a negative control. The numbers show relative protein expression vs. Control (0 mM NaF). Input and Immunoprecipitation (IP) lane relative expression was calculated separately.

Figure 7

Nutlin-3a reversed the fluoride-mediated p21 protein decrease by increasing p-p21 levels. (a) LS8 cells were treated with Nutlin-3a (5 μM) for 2 h followed by the additional NaF (5 mM) for 6 h and then protein was immunoprecipitated using anti-p21 antibody. NaF treatment increased MDM2-p21 binding and Nutlin-3a inhibited this binding. The numbers show relative protein expression vs. control (0 mM NaF). Input and IP lane relative expression was calculated separately for IP lanes. (b) LS8 cells were treated with Nutlin-3a (1–5 μM) for 2 h followed by the additional NaF (5 mM) for 24 h. p21 (18 kDa) and p-p21 (21 kDa) were detected by Western blots. Nutlin-3a treatment (24 h) reversed fluoride-mediated p21 degradation and increased p-p21 levels. The numbers show relative expression normalized by the loading control β-actin (44 kDa). Statistical analysis of relative protein expression of p21 and p-p21 are shown in supplementary Figure S8. (c) Cells were treated with fluoride (5 mM) with/without Nutlin-3a for 24 h. p-p21 (green), nucleus (DAPI; blue) and β-actin (red) expression were detected by immunocytochemistry. Nutlin-3a addition augmented p-p21 expression compared to NaF treatment alone. (d) LS8 cells were treated with Nutlin-3a (5 μM) for 2 h followed by the additional NaF (5 mM) for 24 h. Nutlin-3a alone significantly increased p21 mRNA compared to controls, but Nutlin-3a treatment with NaF did not alter p21 expression compared to NaF alone. Data are presented as the mean ± SD (**; p < 0.01, *; p < 0.05 vs. 0 mM).

Figure 8

Nutlin-3a ameliorated fluoride-induced apoptosis in LS8 cells. LS8 cells were treated with Nutlin-3a (1 μM or 5 μM) for 2 h followed by the addition of NaF (5 mM) for 24 h. DNA damage marker γH2AX (15 kDa) expression and caspase-3 cleavage (17 kDa) were detected by Western blot. Nutlin-3a attenuated caspase-3 cleavage and reduced γH2AX expression. The numbers show relative expression normalized by the loading control β-actin (44 kDa). Statistical analysis of relative protein expression of cleaved-caspase-3 and γH2AX are shown in supplementary Figure S9.

Figure 9

Schema of MDM2-p53 and MDM2-p21 signaling in fluoride toxicity. Fluoride increases acetylated-p53 (Ac-p53) levels to upregulate transcription of Mdm2 and p21. In the early phase (1–6 h), fluoride induces phosphorylation of p21 (p-p21), which translocates p21 from the nucleus to the cytoplasm where p-p21 counteracts fluoride-induced apoptosis. Fluoride enhances MDM2-p53 and MDM2-p21 formation to promote MDM2-mediated p53 and p21 proteasomal degradation that leads to p21 and p-p21 attenuation in the late phase (24 h). Nutlin-3a inhibits MDM2-p53 and MDM2-p21 binding. Nutlin-3a or MG-132 (proteasome inhibitor) reverses fluoride-induced p21 attenuation and increases p-p21 to suppress fluoride-mediated apoptosis.