p53 coordinates with Δ133p53 isoform to promote cell survival under low-level oxidative stress

Dear Editor,

Reactive oxygen species (ROS) can serve as intracellular signals that promote cell proliferation and survival at sub-toxic levels, or function as toxic substances that cause cell death and senescence at high levels (Weinberg and Chandel, 2009). p53 plays a key role in the control of cellular response to ROS by upregulating the expression of either antioxidant genes under low level of oxidative stresses or pro-oxidative and apoptotic genes under high level of stresses (Vigneron and Vousden, 2010; Hafsi and Hainaut, 2011). However, how p53 differentially regulates gene expression in response to different ROS level remains elusive. Δ133p53 is an N-terminal truncated form of p53 (Bourdon et al., 2005) and functions to antagonize p53 apoptotic activity and to promote DNA double-strand break repair (Chen et al., 2009; Aoubala et al., 2011; Gong et al., 2015). In this study, we investigated the functional interaction between p53 and Δ133p53 in response to various levels of ROS.

We treated QSG-7701 cells (a non-cancerous liver epithelial cell line containing WT p53) with increasing concentrations of H₂O₂ (0, 25, 50, 100, 200, and 400 μM) and measured p53 and Δ133p53 expression at 24 h post-treatment (hpt). Interestingly, while protein and transcript expressions of p53 were dose-dependently induced by H₂O₂, in the induction of Δ133p53 expression appeared at the lower dose range, with the peak at 50 μM H₂O₂ (Figure 1A, Supplementary Figure S1A). Upon the exposure to 50 μM H₂O₂, p53 protein started to accumulate at 4 hpt whereas Δ133p53 at 8 hpt (Supplementary Figure S1B). The induction of Δ133p53 by 50 μM H₂O₂ was also observed in other p53-WT cell lines including HepG2, HCT116, and CCD-1079sk, but not in p53-mutated cell lines such as H1299, HCT116 (p53^−/−), and PC-3 (p53 mutant p53^M/M) (Supplementary Figure S1C), confirming that Δ133p53 is a p53 target gene (Marcel et al., 2010).

Figure 1.

Δ133p53 promotes cell survival under sub-toxic level of oxidative stress by upregulating the expression of antioxidant genes. (A) Western blot of human p53 and Δ133p53 at 24 hpt in QSG-7701 cells treated with the indicated concentrations of H₂O₂. Antibody DO-1 for p53, antibody CM1 for Δ133p53. (B) The relative cell viability at 48 hpt of QSG-7701 cells transfected with a non-specific siRNA control (siNS), Δ133p53 siRNA (Δ133p53i), or CMV-Δ133p53 plasmid and then treated with 50 μM H₂O₂, analyzed by MTT assay. (C) FACS analysis of DHE fluorescence levels at 24 hpt in QSG-7701 cells transfected with siNS, Δ133p53i, or CMV-Δ133p53 and then treated with 0 or 50 μM H₂O₂. The relative DHE fluorescence intensity in 50 μM H₂O₂-treated cells is expressed as a fold change against that in the corresponding untreated control cells. (D–F) Relative mRNA expression of SESN1 and SOD1 at 12 hpt in QSG-7701 cells (D and E) or H1299 cells (F) transfected and treated as indicated, measured with qRT-PCR. Gene expression was normalized against β-actin. (G) The SESN1 and SOD1 promoters. The arrows correspond to the orientations of the quarter sites. R = A or G, W = A or T, Y = C or T. (H) ChIP of RE in SESN1 and SOD1 promoters in cells transfected with HA-Δ133p53 or HA-p53, and then treated with 50 μM H₂O₂. The HA antibody was used to co-immunoprecipitate the protein–DNA complex. IgG was used as a non-specific binding control. Specific primer pairs were designed to amplify the corresponding REs. DNA was normalized with a pair of negative control primers for GADPH exon. All statistically significant differences between treatments were calculated from three repeat experiments and assessed with an independent-samples T-test (*P < 0.05, **P < 0.01).

Cell viability was measured to examine the biological significance of H₂O₂-induced Δ133p53. In line with the biphasic effect of ROS on cellular viability, low-level H₂O₂ enhanced whereas high-level H₂O₂ suppressed cell survival (Supplementary Figure S2A–C). Of note, H₂O₂-induced cell survival response mirrored the change in Δ133p53 expression, i.e. the increased cell survival in 50 μM H₂O₂-treated QSG-7701 cells was accompanied with increased Δ133p53 expression. This observation prompted us to assess the role of Δ133p53 in cell survival. By employing a strategy of complementary overexpression and knockdown (RNAi) of Δ133p53, we found that Δ133p53 overexpression augmented whereas knockdown diminished the increased cell viability by 50 μM H₂O₂, demonstrated by MTT, WST-8 assay, and Trypan blue staining (Figure 1B, Supplementary Figure S3A–C). In addition, the fluorescence-activated cell sorting (FACS) analysis of BrdU incorporation revealed that cell viability was highly correlated with the rate of DNA synthesis (Supplementary Figure S3D and E). Similar results were observed in HCT116 cells, a colon cancer cell line containing WT p53 (Supplementary Figure S4A–D).

Next, menadione (vitamin K₃) was used to test whether other oxidants also induced Δ133p53 expression and the effect on cell viability. Similar to H₂O₂, a low concentration of menadione (0.5 μM) resulted in the maximum induction of Δ133p53, which was associated with improved cell survival (Supplementary Figure S5A and B). Similarly, knockdown of Δ133p53 decreased whereas overexpression enhanced cell viability by 0.5 μM menadione (Supplementary Figure S5C and D). These results suggested that Δ133p53 regulates the cell survival response not only to H₂O₂, but also to other oxidants.

H₂O₂ may get converted to free radicals in cells. As measured by dihydroethidium (DHE) staining, the intracellular ${{\text{O}}_2}^{\bullet -}$ level was not changed in control or Δ133p53-overexpressing cells, but significantly increased in Δ133p53-knockdown cells from 12 to 48 hpt of 50 μM H₂O₂ (Figure 1C, Supplementary Figure S6). Treatment with diphenylene iodonium (DPI), an NADPH oxidase inhibitor that blocks the production of ${{\text{O}}_2}^{\bullet -}$, not only prevented the increase of intracellular ${{\text{O}}_2}^{\bullet -}$ (Supplementary Figure S7A and B), but also rescued the decreased cell viability in Δ133p53-knockdown cells (Supplementary Figure S7C). These results implicated an antioxidative function of Δ133p53.

Results from FACS analysis using anti-Annexin V antibody staining (Supplementary Figure S8A and B) and BrdU incorporation analysis (Supplementary Figure S3D and E) suggested that 50 μM H₂O₂ increases cell viability by promoting cell proliferation rather than inhibiting cell death, which is dependent on the expression of Δ133p53. By performing senescence-associated β-galactosidase (SA-β-gal) staining, we showed that knockdown of Δ133p53 significantly augmented and overexpression of Δ133p53 suppressed the cellular senescent response induced by 6-day treatment of 50 μM H₂O₂ (Supplementary Figure S9A and B). To further determine whether Δ133p53 was involved in ROS-induced DNA damage, comet assays were performed at pH7, mainly for detecting DNA double-strand breaks (DSB), and pH10, for detecting both single-strand breaks (SSB) and DSBs, on cells treated with 50 μM H₂O₂ for 6 days. The results showed that only Δ133p53-knockdown cells exhibited significant increase in DNA damage when assayed at pH10 (Supplementary Figure S9C and D), suggesting that Δ133p53 protects cells from low-level H₂O₂-induced senescence by reducing DNA SSBs rather than DNA DSBs. This is consistent with the activity of Δ133p53 to suppress the production of free radicals, confirming its role as an antioxidant.

We then determined the expression level of six p53-regulated antioxidant genes including SESN1, SESN2, SOD1, SOD2, GPX1, and ALDH4A1, together with a p53 response gene p21. Three antioxidant genes (SESN1, SOD1, and SOD2) and p21 were significantly upregulated by 50 μM H₂O₂, which was further enhanced by overexpression of Δ133p53 (Figure 1D, Supplementary Figure S10). Interestingly, knockdown of Δ133p53 compromised the upregulated expression of SESN1 and SOD1 but not that of SOD2 or p21 (Figure 1D, Supplementary Figures S10 and S11). It is also of note that camptothecin, a DNA damage drug, upregulated p21 expression but not SESN1 or SOD1 expression (Supplementary Figure S12).

Furthermore, knockdown or knockout of p53 completely abrogated the induction of SOD1 and SESN1 (Figure 1E and F, Supplementary Figures S13A and S14) and resulted in greater amount of intracellular ${{\text{O}}_2}^{\bullet -}$ and cell death (Supplementary Figure S13B and C) by 50 μM H₂O₂. Notably, in the absence of p53, Δ133p53 lost its ability to induce SOD1 and SESN1 expression, inhibit intracellular ${{\text{O}}_2}^{\bullet }$ accumulation, or increase cell viability, while in the presence of functional p53, expression of Δ133p53 substantially boosted the expression of SOD1 and SESN1 by 50 μM H₂O₂, which was associated with decreased intracellular ${{\text{O}}_2}^{\bullet }$ and cell death (Figure 1E and F, Supplementary Figure S13B and C).

Co-immunoprecipitation assay (Co-IP) demonstrated that Δ133p53 and p53 proteins formed a complex in cells treated with 50 μM H₂O₂ (Supplementary Figure S15), suggesting that Δ133p53 upregulates the expression of antioxidant genes by interacting with p53. Bioinformatic analysis identified the putative p53 response element (RE) within the promoter region of SESN1 (at –1208 nt) and SOD1 (at –858 nt), respectively (Figure 1G) (Gong et al., 2015). Chromatin immuno-precipitation (ChIP) assay revealed that while p53 was enriched at the RE in the promoters of SESN1 and SOD1 in both 50 μM H₂O₂-treated and untreated cells, Δ133p53 was only enriched in 50 μM H₂O₂-treated cells, but not in untreated cells (Figure 1H). This finding suggested that Δ133p53 coordinates with p53 to upregulate the expression of SESN1 and SOD1 upon exposure to 50 μM H₂O₂ by inducible binding to the p53 RE in the promoters of SESN1 and SOD1. Knockdown of SOD1 or SESN1 expression abolished the ability of Δ133p53 to enhance cell survival in response to the treatment with 50 μM H₂O₂ (Supplementary Figure S16A and B), demonstrating that either SOD1 or SESN1 was required for Δ133p53-dependent promotion of cell survival.

Either excess amount of ROS or decreased cellular antioxidant capacity leads to aging and human diseases by causing oxidative damages. Our study demonstrates that Δ133p53 promotes cell survival in response to sub-toxic level of oxidative stress by coordinating with p53 to upregulate the expression of antioxidant genes such as SOD1 and SESN1, and thus may play an important role in oxidative stress-induced aging process and human pathologies.

[Supplementary material is available at Journal of Molecular Cell Biology online. We thank Professors Binghui Shen, Yingjie Wang, and Wen Yi for sharing lab equipment and materials. This work was supported by the International Science and Technology Cooperation Program of China (2013DFG32910), the ‘973 Program’ (2012CB944500), the National Natural Science Foundation of China (31371491 and 30971677), and Zhejiang Provincial Natural Science Foundation of China (No. LZ13C120001). L.G. and Z.-M.Y. were supported by NIH/NCI (R01 CA85679).]