Role of Stat3 in regulating p53 expression and function

Abstract

Loss of p53 function by mutation is common in cancer. However, most natural p53 mutations occur at a late stage in tumor development, and many clinically detectable cancers have reduced p53 expression but no p53 mutations. It remains to be fully determined what mechanisms disable p53 during malignant initiation and in cancers without mutations that directly affect p53. We show here that oncogenic signaling pathways inhibit the p53 gene transcription rate through a mechanism involving Stat3, which binds to the p53 promoter in vitro and in vivo. Site-specific mutation of a Stat3 DNA-binding site in the p53 promoter partially abrogates Stat3-induced inhibition. Stat3 activity also influences p53 response genes and affects UV-induced cell growth arrest in normal cells. Furthermore, blocking Stat3 in cancer cells up-regulates expression of p53, leading to p53-mediated tumor cell apoptosis. As a point of convergence for many oncogenic signaling pathways, Stat3 is constitutively activated at high frequency in a wide diversity of cancers and is a promising molecular target for cancer therapy. Thus, repression of p53 expression by Stat3 is likely to have an important role in development of tumors, and targeting Stat3 represents a novel therapeutic approach for p53 reactivation in many cancers lacking p53 mutations.

FIG. 1.

p53 expression is inhibited at both RNA and protein levels in a Stat3-dependent manner. (A) p53 mRNA and protein levels in BALB/c 3T3 and v-Src 3T3 fibroblasts were compared by Northern blot and Western blot analyses. GAPDH and β-actin were used here to ensure that similar amounts of total RNAs and proteins were loaded in each lane. (B) Transient transfection of v-Src expression vector enhances Stat3 activity (EMSA, top panel) and inhibits p53 RNA levels (Northern blot analysis, lower panel). Cotransfecting v-Src expression vector with a Stat3 expression vector further inhibits p53 expression. (C) p53 expression is regulated by Stat3. Top gel, Stat3 DNA-binding activity in the indicated cells as measured by EMSA. Middle gel, Northern blot analysis. Bottom gel, Western blot analysis. 3T3, 3T3 BALB/c fibroblasts; v-Src, v-Src 3T3 fibroblasts; MEFs, mouse embryonic fibroblasts. Stat3D, 3T3 cells stably transduced with MSCV-Stat3D; Stat3C, MEFs transfected with Stat3C plasmid expression vector. wt, wild type.

FIG. 2.

Stat3 activity inhibits p53 transcription. (A) 3T3 fibroblasts were transfected with a p53 promoter/luciferase reporter construct in the presence of a control empty vector, pcDNA3, a wild-type Stat3 expression vector, a constitutively activated c-Src mutant, c-Src531, or both wild-type (WT) Stat3 and a c-Src mutant as indicated. Luciferase activity was normalized to transfection efficiency using β-galactosidase activity as an internal control. Luciferase activity of pcDNA3-transfected cells was assigned as 100%. Data represent means ± standard deviations; n = 3. (B) Constitutive Stat3 activity inhibits p53 promoter activity. 3T3 fibroblasts were transfected with either a control empty vector, pcDNA3, or a Stat3C expression vector in the presence of one of the promoter/luciferase reporter constructs as indicated. Luciferase activity was normalized to transfection efficiency using β-galactosidase activity as an internal control. (C and D) Nuclear run-on assays. ³²P-labeled nuclear RNAs were hybridized to membranes containing p53 and GAPDH cDNA fragments. Stat3 activation was induced by either PDGF (10 ng/ml for 8 h) or v-Src. Stat3D was used to block Stat3 signaling in v-Src-transformed 3T3 cells. These experiments were repeated twice with similar results. The p53/GAPDH ratio was based on band intensities determined by phosphorimaging.

FIG. 3.

Stat3 protein binds to the p53 promoter in vivo (ChIP assays). (A) Gel electrophoresis of PCR products using primers detecting the p53 promoter following ChIP with the indicated antibodies. Anti-CD40 rabbit polyclonal antibody was used as an irrelevant control antibody. In the case of the Stat1 antibody ChIP assay, IFN-γ (2 h at 100 U/ml) was used to stimulate Stat1 activity, which was detected by EMSA (data not shown). These experiments were repeated twice with similar results. (B) PDGF stimulation (10 ng/ml for 8 h), which activates Stat3, also leads to Stat3 binding to the p53 promoter region as shown by ChIP. (C) ChIP assays of chromatin prepared from 3T3 and v-Src 3T3 cells using the indicated antibodies followed by real-time PCR and normalized to the amount of nonspecifically precipitated β-actin promoter in the same samples. Ab, antibody; IgG, immunoglobulin G.

FIG. 4.

Stat3 protein binds to the p53 promoter and contributes to repression of the p53 promoter transcription activity. (A) Stat3 protein and the putative Stat3 DNA-binding site in the p53 promoter form protein-DNA complexes (EMSA). Top panel, labeled wild-type p53(−218) was able to form a DNA/protein complex (shift), while the mutated p53(−218M) oligonucleotides could not in v-Src 3T3 cells. Stat3 antibody (α-Stat3) but not Stat1 antibody (α-Stat1) was able to supershift the complex. Both wild-type p53(−218) and hSIE oligonucleotides disrupted the DNA/protein complex (shift), but the p53(-218M) mutant and FIRE, a nonspecific oligonucleotide, failed. In addition, labeled p53(218M) (mutant) could not form a DNA/protein complex. Parallel EMSA experiments were performed with a labeled hSIE probe (lower panel), confirming that the p53(-218) Stat3 site can interact with Stat3 protein. (B) Stat3-induced repression of p53 is partially mediated by STAT DNA-binding site at −218. 3T3 fibroblasts were transfected with either pcDNA3 or a Stat3C expression vector in the presence of one of the promoter/luciferase reporter constructs as indicated. An SV40 promoter/luciferase reporter construct was used as a negative control. Data are presented as fold repression by Stat3C (value of luciferase activity in pcDNA3-transfected cells divided by value of luciferase activity in Stat3C-transfected cells; fold repression of SV40 promoter activity by Stat3C was designated as 1). Luciferase activity was normalized to transfection efficiency using β-galactosidase activity as an internal control. Data shown are means ± standard deviations (n = 3). M, mutant.

FIG. 5.

Stat3 activity inhibits p53-responsive promoter activity and UV-induced cell growth arrest. (A) Stat3C inhibits activity of the p53-responsive element mediated by endogenous p53 protein. pGL2-BP100 is a derivative of pGL2-basic promoter/luciferase reporter constructs driven by the p53-responsive element. An expression vector encoding MDM2 was cotransfected into 3T3MSCV and 3T3Stat3C expressing stable cell lines. (B) Constitutive Stat3 activity resists UV-induced growth arrest. For the top panel, 8 h after UV irradiation at indicated UV doses, ³H-TdR incorporation and Western blot analyses for p53 protein levels in the treated 3T3 cells were performed. These experiments were repeated at least twice with similar results. (C) Lack of Stat3 activity in Stat3^−/− MEFs allows an increase in UV-induced p53 expression and enhanced growth inhibition compared to Stat3^+/+ MEFs. (D) Stat3 blockade-mediated, UV-induced decrease in thymidine incorporation is p53 dependent. Stat3^−/− MEFs used in panel C were transiently transfected with control or p53 siRNAs, followed by UV treatment at the indicated doses. The effectiveness of the siRNA was shown by Western blot analysis (lower panel). Data shown represents means ± standard deviations of four independent experiments. P < 0.01 (comparing Stat3^−/− with Stat3^+/+ MEFs as well as comparing control siRNA with p53 siRNA in Stat3^−/− MEFs).

FIG. 6.

Blocking Stat3 in human tumor cells induces p53-dependent tumor cell apoptosis and UV-induced growth arrest. (A) Inhibiting Stat3 activity up-regulates p53 and p21 expression. In the left panel, Western blots show expression of dominant-negative Stat3, Stat3β, in the human melanoma cell line A2058. Overexpression of Stat3β is expected to inhibit wild-type Stat3 activity. The middle panel shows a RNase protection assay of A2058 human melanoma cells. The right panel shows Western blot analysis of stable clones of A2058 tumor cells transfected with either control vector or Stat3 siRNA expressing vector. Fold changes were based on band intensities quantified by phosphorimaging. (B) Blocking Stat3 in tumor cells induces p53-dependent apoptosis. pIRES-EGFP and pIRES-Stat3β were cotransfected with either pcDNA3 control empty vector or the p53 dominant-negative mutant, pCMV-DD. A2058 tumor cells positive for green fluorescent protein (GFP) (transfected cells) were assayed for annexin V binding by flow cytometry. (C) Targeting Stat3 induces apoptosis of p53^+/+ but not p53^−/− HCT116 colon cancer cells. Left panel, Stat3 EMSA indicates Stat3 is activated in HCT116 cells. Right panel, flow cytometry analyses of Annexin V binding of GFP-positive cells; n = 3. (D) Blocking Stat3 in A2058 human melanoma cells allows more UV-induced growth arrest, which is p53 dependent. Left panel, Stat3 siRNA A2058 tumor cells (same as in panel A) were transfected with a control empty vector or a p53 dominant-negative expression vector, pCMV-DD (n = 3). Right panel, Stat3 siRNA A2058 tumor cells (same as in panel A) were transfected with 10 nM of either a control siRNA or p53 siRNA. The effectiveness of the siRNA was shown by Western blot analysis (lower panel). The figure shown represents means ± standard deviations of four experiments. Comparing Stat3 siRNA with control siRNA in A2058, P is <0.01 for both panels.