Double-stranded-RNA-activated protein kinase PKR enhances transcriptional activation by tumor suppressor p53

Abstract

The tumor suppressor p53 plays a key role in inducing G1 arrest and apoptosis following DNA damage. The double-stranded-RNA-activated protein PKR is a serine/threonine interferon (IFN)-inducible kinase which plays an important role in regulation of gene expression at both transcriptional and translational levels. Since a cross talk between IFN-inducible proteins and p53 had already been established, we investigated whether and how p53 function was modulated by PKR. We analyzed p53 function in several cell lines derived from PKR+/+ and PKR-/- mouse embryonic fibroblasts (MEFs) after transfection with the temperature-sensitive (ts) mutant of mouse p53 [p53(Val135)]. Here we report that transactivation of transcription by p53 and G0/G1 arrest were impaired in PKR-/- cells upon conditions that ts p53 acquired a wild-type conformation. Phosphorylation of mouse p53 on Ser18 was defective in PKR-/- cells, consistent with an impaired transcriptional induction of the p53-inducible genes encoding p21(WAF/Cip1) and Mdm2. In addition, Ser18 phosphorylation and transcriptional activation by mouse p53 were diminished in PKR-/- cells after DNA damage induced by the anticancer drug adriamycin or gamma radiation but not by UV radiation. Furthermore, the specific phosphatidylinositol-3 (PI-3) kinase inhibitor LY294002 inhibited the induction of phosphorylation of Ser18 of p53 by adriamycin to a higher degree in PKR+/+ cells than in PKR-/- cells. These novel findings suggest that PKR enhances p53 transcriptional function and implicate PKR in cell signaling elicited by a specific type of DNA damage that leads to p53 phosphorylation, possibly through a PI-3 kinase pathway.

FIG. 1

Generation of PKR^+/+ and PKR^−/− cell lines expressing the ts p53. (A) Northern blot analysis. Total RNA (10 μg) from cell lines transfected with the ts p53 cDNA was subjected to Northern blotting using as a probe (top panel) the ³²P-labeled mouse p53 cDNA, which recognizes both endogenous mouse p53 and ts mouse p53 RNAs. For normalization, the membrane was stripped and reprobed with ³²P-labeled actin cDNA (bottom panel). (B) Immunodetection of p53 in PKR^+/+ and PKR^−/− cells transfected with p53(Val135). Lysates (50 μg of protein) from each clone were resolved by SDS-PAGE (10% gel) and then electrophoretically transferred to a PVDF membrane. p53 protein levels were detected by immunoblotting with PAb421 (top panel), which cross-reacts with both the endogenous mouse p53 and transfected ts p53. The membrane was stripped and reprobed with an antiactin antibody to ensure equal protein loading of the samples (bottom panel). (A and B) Lanes 1 and 4, PKR^+/+ and PKR^−/− cells transfected with the zeocin resistance gene only (control [CON] clones); lanes 2 and 3, PKR^+/+ clones transfected with the ts p53 gene; lanes 5 to 8, PKR^−/− clones transfected with ts p53 gene.

FIG. 2

Impaired expression of p21^WAF1/Cip1 and Mdm2 in PKR^−/− cells expressing p53(Val135). PKR^+/+ (lanes 1 to 6) and PKR^−/− clones (lanes 7 to 12) expressing p53(Val135) (lanes 3 to 6 and 9 to 12, respectively) were maintained for 12 h in conditions such that p53 acquires a mutant (37°C; lanes 1, 3, 5, 7, 9, and 11) or wild-type (32°C; lanes 2, 4, 6, 8, 10, and 12) form. Protein extracts (50 μg) were analyzed for p53, p21, and Mdm2 protein expression by immunoblot analysis using anti-p53 MAb PAb 421 (top panel), a mix of anti-Mdm2 MAbs 2A10 and 4B11 (second panel from top), rabbit antisera to p21 (third panel from top), or antiactin MAb (bottom panel). As controls (CON), PKR^+/+ (lanes 1 and 2) and PKR^−/− (lanes 7 and 8) clones expressing the zeocin resistance gene only were used. To avoid redundancy, results for two of the four PKR^−/− clones overexpressing p53(Val135) are shown.

FIG. 3

PKR is implicated in G₀/G₁ arrest by ts p53. Shown is cell cycle analysis of PKR^+/+ and PKR^−/− clones expressing ts p53. Cells were incubated in DMEM–0.5% FCS for 48 h, refed with 10% FCS, and incubated at 32°C for the indicated periods of time. Cells were then fixed, stained with propidium iodide, and subjected to cell cycle analysis as described in Materials and Methods. M1, fraction of cells in G₀/G₁ phase; M2, cells in S phase; M3, cells in G₂+M.

FIG. 4

Transcriptional activation of p21 and mdm2 genes by p53 is enhanced by PKR. (A and B) PKR^+/+ and PKR^−/− cells expressing ts p53 were either maintained at 37°C (lanes 1 and 6) or switched to 32°C for 1 (lanes 2 and 7), 3 (lanes 3 and 8), 6 (lanes 4 and 9), and 12 (lanes 5 and 10) h. (A) Northern blot analysis. Total RNA (10 μg) was subjected to Northern blot analysis using a mix of ³²P-labeled mdm2 and mouse p21 cDNAs as a probe (top panel). The levels of mdm2 and p21 RNA expression were normalized to actin RNA (bottom panel). (B) Immunodetection of p21 levels in PKR^+/+ and PKR^−/− cells expressing ts p53. Cell extracts (50 μg of protein) were subjected to SDS-PAGE (12% gel), electrotransferred to a PVDF membrane, and probed with an anti-p21 antibody. The blot was stripped and reprobed with an antiactin antibody for normalization of protein extracts. (A and B) Quantification of bands was performed by scanning autoradiograms in the linear range of exposure with a Chemiimager 4000 imaging system and analyzing with spot densitometry software (Alpha Innotech Corporation).

FIG. 5

DNA binding of ts p53 is not affected by PKR. EMSA was performed with cell extracts from control (CON) zeocin-resistant PKR^+/+ and PKR^−/− cells (lanes 1 to 3 and 7 to 9) or cells expressing ts p53 (lanes 4 to 6 and 10 to 12) which were incubated at either 37°C (lanes 1, 4, 7, and 10) or 32°C (lanes 2, 3, 5, 6, 8, 9, 11, and 12). Cell extracts (10 μg of protein) were incubated with a ³²P-labeled dsDNA oligonucleotide which contains the consensus DNA-binding site of p53. Binding reaction mixtures contained 100 ng of affinity-purified anti-p53 MAb PAb421, which induces the sequence-specific DNA binding of p53. A 200-fold excess of unlabeled dsDNA oligonucleotide was added in cold competition reactions (lanes 3, 6, 9, and 12).

FIG. 6

PKR enhances the induction of p21^WAF1 expression upon DNA damage. (A and B) Induction of p21 in PKR^+/+ and PKR^−/− cells in response to adriamycin (AD). Exponentially grown cells (∼3 × 10⁶/150-mm-diameter dish) were left untreated (lanes 1 and 6) or treated with 1 μM adriamycin for 3 (lanes 2 and 7), 6 (lanes 3 and 8), 12 (lanes 4 and 9), and 24 (lanes 5 and 10) h. Protein extracts (∼250 μg) were subjected to immunoprecipitation with anti-p53 antibody PAb421, transferred onto a PVDF membrane, and immunoblotted with the rabbit antiserum to p53 (CM-1; A, top panel). Cell extracts (50 μg of protein) were resolved by SDS-PAGE (12% gel) and subjected to immunoblot analysis first with anti-p21 antibody (A, middle panel; B, top panel) and then with antiactin antibody (bottom panels). (A) Spontaneously immortalized polyclonal populations derived from PKR^+/+ and PKR^−/− MEFs. (B) Primary PKR^+/+ and PKR^−/− MEFs grown between passages 2 and 3. (A and B) Quantification of bands performed by scanning autoradiograms in the linear range of exposure with a Chemiimager 4000 imaging system and analyzing the results with spot densitometry software (Alpha Innotech Corporation).

FIG. 7

PKR enhances phosphorylation of ts p53 on Ser¹⁸. (A) The anti-P-Ser¹⁵ human p53 antibody specifically recognizes P-Ser¹⁸ of mouse p53. BALB/c(10)1 cells lacking endogenous p53 were transiently transfected either with a wild-type mouse p53 cDNA (lanes 3 and 4) or a Ser¹⁸-Ala mutant (mut) of mouse p53 cDNA (lanes 5 and 6). Thirty-six hours after transfection, cells were treated with adriamycin (AD) for 1 h (lanes 2, 4, and 6). Cell extracts (50 μg of protein) were subjected to SDS-PAGE (12% gel), electrotransferred to a PVDF membrane, and probed first with the anti-P-Ser¹⁵ human p53 antibody (top panel) and second with mouse anti-p53 antibody PAb421 (bottom panel). Lanes 1 and 2, mock-transfected cells before (lane 1) and after (lane 2) adriamycin treatment. (B) Immunodetection of ts p53 phosphorylation in PKR^+/+ and PKR^−/− cells, using a mouse MAb specific for P-Ser¹⁵ of human p53. PKR^+/+ and PKR^−/− cells expressing ts p53 were either maintained at 37°C (lanes 1 and 4) or shifted to 32°C for 3 (lanes 2 and 5) and 6 (lanes 3 and 6) h. (C) For DNA damage, cells were shifted to 32°C for 6 h (lanes 1 and 4) and preincubated with the proteasome inhibitors ALLN and MG-132 for 1 h (lanes 1 to 6) before treatment with adriamycin for 3 (lanes 2 and 5) or 6 (lanes 3 and 6) h. Cell extracts (50 μg of protein) were subjected to SDS-PAGE (10% gel) and immunoblot analysis first with the mouse anti-P-Ser¹⁵ MAb (B and C, top panels) and then with mouse anti-p53 MAb PAb421 (bottom panels).

FIG. 8

PKR enhances p53 phosphorylation on Ser¹⁸ in response to DNA damage. Cell extracts from PKR^+/+ and PKR^−/− cells were prepared after treatment with adriamycin (AD) (A), γ radiation (7 Gy), (B), or UV radiation (50 J/m²) (C) for the indicated periods of time postinfection (P.I.). Protein extracts (∼250 μg) were subjected to immunoprecipitation with a rabbit anti-p53 polyclonal antibody, transferred onto a PVDF membrane, and immunoblotted with the mouse anti-P-Ser¹⁵ MAb (upper panels). The membranes were stripped and reprobed with the mouse anti-p53 antibody PAb421 (lower panels). The faint slowly migrating band above p53 present in all lanes after immunoblotting with the anti-P-Ser¹⁸ antibody (B and C, top panels) is a fraction of the heavy chain of rabbit IgG (IgG H.C.) recognized by the secondary HRP-conjugated sheep anti-mouse IgG and was visible after long exposures only.

FIG. 9

PKR is implicated in PI-3 kinase pathway(s) leading to p53 phosphorylation. PKR^+/+ and PKR^−/− cells were treated with 10 μM specific PI-3 kinase inhibitor LY294002 and 1 μM adriamycin (AD) in the presence of ALLN and MG-132. To avoid a possible degradation of LY294002 during the treatment, adriamycin was added for 1 h only. Protein extracts (∼250 μg) were subjected to immunoprecipitation with a rabbit anti-p53 polyclonal antibody and resolved by SDS-PAGE (10% gel) followed by immunoblotting with the mouse anti-P-Ser¹⁵ MAb (upper panels). The membranes were stripped, and p53 protein levels were visualized by immunoblot analysis using mouse anti-p53 antibody PAb421 (lower panels). The faint slowly migrating band above p53 (top panel, lanes 5 to 8) is a fraction of the heavy chain of rabbit IgG (IgG H.C.) recognized by the secondary HRP-conjugated sheep anti-mouse IgG.