Expression of p53 enhances selenite-induced superoxide production and apoptosis in human prostate cancer cells

Abstract

Although the anticancer effects of selenium have been shown in clinical, preclinical, and laboratory studies, the underlying mechanism(s) remains unclear. Our previous study showed that sodium selenite induced LNCaP human prostate cancer cell apoptosis in association with production of reactive oxygen species, alteration of cell redox state, and mitochondrial damage. In the present study, we showed that selenite-induced apoptosis was superoxide mediated and p53 dependent via mitochondrial pathways. In addition, we also showed that superoxide production by selenite was p53 dependent. Our study showed that wild-type p53-expressing LNCaP cells were more sensitive to selenite-induced apoptosis than p53-null PC3 cells. Selenite treatment resulted in high levels of superoxide production in LNCaP cells but only low levels in PC3 cells. LNCaP cells also showed sequential increases in levels of phosphorylated p53 (serine 15), total p53, Bax, and p21(Waf1) proteins following selenite treatment. The effects of selenite were suppressed by pretreatment with a synthetic superoxide dismutase mimic or by knockdown of p53 via RNA interference. LNCaP cells treated with selenite also showed p53 translocation to mitochondria, cytochrome c release into the cytosol, and activation of caspase-9. On the other hand, restoration of wild-type p53 expression in PC3 cells increased cellular sensitivity to selenite and resulted in increased superoxide production, caspase-9 activation, and apoptosis following selenite treatment. These results suggest that selenite induces apoptosis by producing superoxide to activate p53 and to induce p53 mitochondrial translocation. Activation of p53 in turn synergistically enhances superoxide production and apoptosis induced by selenite.

Figure 1

Selenite causes apoptosis with superoxide accumulation in wt-p53 expressing LNCaP cells. A, Dose-dependent effect of selenite on cell viability as demonstrated by the MTT assay. Cells were treated with 0-3.5 μM selenite for 5 days. B, Agarose gel electrophoretic detection of DNA fragmentation as a marker of cell apoptosis induced by selenite. Cells were treated with 0–2.5 μM selenite for 24 hr. C, Flow cytometry analyzed cell apoptosis by measuring the sub-G1 cell population. Cells were treated with 2.5 μM selenite for 24 and 48 hr. D, The SOD mimic MnTMPyP protected against cytotoxicity of selenite. Cells were treated with 0–3.5 μM selenite with or without 5 μM MnTMPyP for 5 days. E, Chemiluminescence assay showing production of superoxide radicals in cells treated with selenite. Cells were treated with 2.5 μM selenite, 5 μM MnTMPyP or selenite plus MnTMPyP. The treatment agents were added into the cell suspensions in test tubes and chemiluminescence was immediately measured. Data are presented as means ± SD of three independent experiments. * p<0.05 compared with 0 μM; # p<0.05 compared with the corresponding concentration of selenite without MnTMPyP; ** p<0.05 compared with control (cells without treatment), MnTMPyP, Se + MnTMPyP, and Se at 1 min.

Figure 2

Western blot analysis of effects of selenite on expression of p53, p21^Waf1, and Bax and phosphorylation of p53 and histone (H2AX) in LNCaP cells. A, Dose-dependent effect of selenite. Cells were treated with 0–3.5 μM selenite for 18 hr. B, Time-dependent effect of selenite. Cells were treated with 2.5 μM selenite for 0–36 hr. C, The SOD mimic MnTMPyP suppressed selenite effects on p53, p21^Waf1, and Bax. Cells were treated with 2.5 μM selenite, 5 μM MnTMPyP, or combination of selenite and MnTMPyP for 18 hr. Protein loading: 40 μg for p53, p-p53 Ser15, p21^Waf1, Bax, and H2AX and 20 μg for β-actin.

Figure 3

Downregulation of p53, but not Bax, by RNA interference causes resistance to selenite-mediated cytotoxicity in LNCaP cells. A, MTT assay of viability of LNCaP cells with p53 siRNA transfection and selenite treatment. B, Western blot analysis showing suppressive effects of p53 siRNA transfection on expression of p53, p21^Waf1, and Bax in cells with and without selenite treatment. C, MTT assay of cellular response to selenite following downregulation of Bax by siRNA transfection. D, Western blot analysis of effects of Bax siRNA transfection on p53, p21^Waf1, and Bax in cells with and without selenite treatment. Cells were transfected with 50 nM of p53 or Bax siRNA, respectively, for 36 hr and then treated with selenite for 5 days for cell viability analysis or with 2.5 μM selenite for 18 hr for Western blot analysis. Protein loading: 40 μg for p53, p-p53 Ser15, p21^Waf1, Bax, and H2AX and 20 μg for β-actin. The data were obtained from three independent experiments and the results shown are the mean ± SD. *p<0.05 compared with RNAiFect and negative control siRNA.

Figure 4

Forced p53 expression sensitizes PC3 cells to selenite-mediated cytotoxicity. A, MTT assay of a dose-dependent effect of selenite on cell viability. Cells were treated with selenite for 5 days. B, MTT assay of a time-dependent effect of selenite on cell viability. Cells were treated with 2.5 μM selenite. C, Western blot analysis of levels of p53, p-p53 Ser 15, p21^Waf1, and Bax in cells following Ad-p53 transfection and selenite treatment. D, MTT assay of effect of p53 on viability of cells with or without selenite treatment. Cells were transduced with Ad-p53 constructs for 36 hr and then treated with 2.5 μM selenite for 18 hr for western blot analysis and for 5 days for cell viability analysis. The data were obtained from three independent experiments and the results shown are the mean± SD. *p<0.05 compared with 0 μM selenite or 0 hr. ** p<0.05 compared with control and 4 MOI Ad.

Figure 5

p53 modulates selenite-induced production of superoxide radical in LNCaP and PC3 cells. A, Chemiluminescence assay of suppression of superoxide production by p53 siRNA transfection in LNCaP cells treated with selenite. B, Enhancement of superoxide production by Ad-p53 transduction in PC3 cells treated with selenite. Cells were transfected with 50 nM p53 siRNA or transduced with 4 MOI Ad-p53 for 36 hr and then treated with 2.5 μM selenite in suspension. Superoxide production was immediately measured using a luminometer. The data were obtained from three independent experiments and the results shown are the mean± SD. *p<0.05 control (cells without treatment) vs. p53 siRNA and p53 siRNA + Se. ** p<0.05 2.5 μM Se vs. control, p53 siRNA, and p53 siRNA + Se. ^#p<0.05 control (cells without treatment) vs. 2.5 μM Se and Ad-p53. ^##p<0.05 Ad-p53 + Se vs. control, 2.5 μM Se, and Ad-p53.

Figure 6

Selenite causes p53 mitochondrial translocation, cytochrome c release from mitochondria, and activation of caspase 9 in LNCaP and PC3 cells. A, Western blot analysis of selenite-induced p53 accumulation in mitochondria and release of cytochrome c into cytosol in LNCaP cells. WC, whole cell lysate; Cyto, cytosol extract; and Mito, mitochondrial extract. Cells were treated with 2.5 μM selenite for 18 hr. p21^Waf1 was used as a control for the purity of mitochondrial extracts and also for p53 transcriptional activity. β-actin was used as a control for equal sample loading. Protein loading: 40 μg for p53, p-p53 Ser15, p21^Waf1, Bax, and cytochrome c, and 20 μg for β-actin. B, Laser scanning confocal microscopic photographs showing co-localization of p53 and MitoTracker in mitochondria of LNCaP cells. Red, MitoTracker as a mitochondrial maker; green, p53 labeled with a p53 antibody and a FITC-conjugated secondary antibody; overlay, merging of MitoTracker (red) and p53 (green). C, Chemiluminescence assay of activation of caspase 9 by selenite treatment and the suppressive effect of p53 siRNA transfection in LNCaP cells. D, Chemiluminescence assay of activation of caspase 9 by selenite treatment and the enhancing effect of Ad-p53 transduction in PC3 cells. Cells were transfected with 5 nM p53 siRNAs or transduced with 4 MOI p53 Ad for 36 hr and then treated with 2.5 μM selenite for 18 hr. Chemiluminescence was measured using a luminometer. The data were obtained from three independent experiments and the results shown are the mean± SD. *p<0.05 compared with control (cells without treatment), p53 siRNA, and p53 siRNA+Se or control (LNCaP), Se, and Ad-p53 (PC3).