Translational repression of p53 by RNPC1, a p53 target overexpressed in lymphomas

Abstract

The p53 pathway is critical for tumor suppression, as the majority of human cancer has a faulty p53. Here, we identified RNPC1, a p53 target and a RNA-binding protein, as a critical regulator of p53 translation. We showed that ectopic expression of RNPC1 inhibited, whereas knockdown of RNPC1 increased, p53 translation under normal and stress conditions. We also showed that RNPC1 prevented cap-binding protein eIF4E from binding p53 mRNA via its C-terminal domain for physical interaction with eIF4E, and its N-terminal domain for binding p53 mRNA. Consistent with this, we found that RNPC1 directly binds to p53 5' and 3'untranslated regions (UTRs). Importantly, we showed that RNPC1 inhibits ectopic expression of p53 in a dose-dependent manner via p53 5' or 3' UTR. Moreover, we showed that loss of RNPC1 in mouse embryonic fibroblasts increased the level of p53 protein, leading to enhanced premature senescence in a p53-dependent manner. Finally, to explore the clinical relevance of our finding, we showed that RNPC1 was frequently overexpressed in dog lymphomas, most of which were accompanied by decreased expression of wild-type p53. Together, we identified a novel p53-RNPC1 autoregulatory loop, and our findings suggest that RNPC1 plays a role in tumorigenesis by repressing p53 translation.

Figure 1.

p53 expression is inhibited by ectopic expression of RNPC1a but increased by knockdown or knockout of RNPC1 under normal and stress conditions. (A) Schematic illustration of the RNPC1 locus and the usage of exons for RNPC1a and RNPC1b. (B) RNPC1a inhibits p53 expression. Western blots were prepared with extracts from MCF7, RKO, and HCT116 cells uninduced or induced to express HA-tagged RNPC1a for 24 h and probed with antibodies against HA, p53, or actin. The basal levels of p53 were arbitrarily set at 1.0 and the fold change is shown below each lane. (C) The levels of RNPC1a, p53, MDM2, ECT2, and GAPDH were measured in MCF7 or RKO cells uninduced or induced to express RNPC1a for 12 h, followed by mock treatment or treatment with nutlin-3 or doxorubicin for 12 h. (D) The basal level of p53 is increased by total RNPC1 or RNPC1a knockdown. MCF7, RKO, and HCT116 cells were transiently transfected with scrambled siRNA or siRNA against RNPC1a or total RNPC1 for 3 d, and the levels of RNPC1a, p53, and actin were analyzed by Western blot analysis. (E) The level of mutant p53 is increased by RNPC1 knockdown in SW480 cells. The experiments were performed as in C. (F) The levels of RNPC1a, p53, and actin were measured in MCF7 cells uninduced or induced to express shRNA against total RNPC1 for 3 d. (G) The levels of RNPC1a, p53, and actin were measured in RKO cells transfected with scrambled siRNA or siRNA against RNPC1a or total RNPC1 for 3 d, followed by treatment with or without doxorubicin for 12 h. (H) The levels of RNPC1a, p53, and actin were measured in MCF7 cells uninduced or induced to express shRNA against total RNPC1 for 3 d, followed by mock or doxorubicin or camptothecin treatment for 12 h. (I) MEFs isolated from wild-type or RNPC1^−/− embryos at passage 3 were treated with or without doxorubicin for 12 h, and the levels of RNPC1, p53, and GAPDH protein were analyzed by Western blot analysis.

Figure 2.

p53 protein translation is inhibited by ectopic expression of RNPC1a but increased by knockdown of RNPC1 under normal and stress conditions. (A) The level of newly synthesized p53 protein is decreased by RNPC1a. HCT116 and RKO cells were uninduced or induced to express RNPC1a for 12 h and then ³⁵S-labeled for 5 min, followed by immunoprecipitation with anti-p53. The immunocomplexes were resolved by SDS-PAGE, and p53 was visualized by autoradiography. The input was blotted with anti-RNPC1 to ensure RNPC1a expression, shown below each column. (B) The level of newly synthesized p53 protein is decreased by RNPC1a in SW480 cells. The experiment was performed as in A, except that SW480 cells were used and ³⁵S-labled for 20 min. (C) The level of newly synthesized p53 protein is increased by RNPC1 knockdown. The experiment was performed as in A with MCF7 cells uninduced or induced to knock down RNPC1a for 3 d and then ³⁵S-labeled for 5 min. (D) p53 translation is increased upon DNA damage. RKO cells were treated with or without doxorubicin or camptothecin for 24 h, and the level of newly synthesized p53 protein was measured by ³⁵S-labeling. (E) RNPC1a inhibits p53 translation under a stress condition. RKO cells were uninduced or induced to express RNPC1a for 12 h, followed by treatment with doxorubicin for 24 h, and the level of newly synthesized p53 protein was measured by ³⁵S-labeling. (F) Schematic presentation of free, monoriobsomal, and polysomal mRNA distribution after sucrose gradient sedimentation. (G) RNPC1a inhibits the association of heavy polysomes with p53 mRNA. Polysomes were separated by sucrose density gradient from HCT116 cells uninduced or induced to express RNPC1a for 12 h. p53 and actin transcripts were detected in each fraction by RT–PCR. (H) The experiment was performed as in D, except that cell extracts were treated with puromycin.

Figure 3.

RNPC1a prevents eIF4E from binding p53 mRNA, which is dependent on its N-terminal RRM for binding p53 mRNA. (A) The level of eIF4E, RNPC1, p53, and actin was measured in HCT116 and RKO cells transfected with scrambled siRNA or siRNA against eIF4E for 3 d, followed with or without RNPC1a induction for 12 h. (B) The experiment was performed as in A, except that MCF7 cells were transfected with scrambled siRNA or siRNA against eIF4E along with or without RNPC1 knockdown for 3 d. (C,D) RNPC1a prevents eIF4E from binding p53 mRNA. RKO (C) and HCT116 (D) cells were uninduced or induced to express RNPC1a for 18 h, followed by immunoprecipitation with antibodies against HA-tagged RNPC1a or eIF4E, or with mouse IgG. Total RNAs were purified from immunocomplexes and subjected to RT–PCR to measure the level of p53 and actin mRNAs associated with RNPC1a or eIF4E. (E) The association of p53 and actin transcripts with RNPC1a or eIF4E is RNA-dependent. The same experiment was performed as described in C, except that cell lysates were treated with RNase prior to immunoprecipitation. (F) RNPC1b is unable to prevent eIF4E from binding p53 mRNA. The experiment was performed as described in C with RKO cells uninduced or induced to express RNPC1b. (G,H) RNPC1a is able to bind to p53 mRNA regardless of eIF4E knockdown. The experiment was performed as described in C, except that RKO cells were transfected with scrambled siRNA (G) or siRNA against eIF4E (H) for 3 d, followed with or without RNPC1a induction for 12 h. (I) Schematic illustration of RNP1 and RNP2 deletion mutants. (J,K) The RNA-binding domain is required for RNPC1 to prevent eIF4E from binding p53 mRNA. The experiment was performed as in C with RKO cells uninduced or induced to express ΔRNP1 (J) or ΔRNP2 (K). (L) ΔRNP1 and ΔRNP2 are unable to inhibit p53 expression. RNPC1, p53, and GAPDH were examined in RKO cells transfected with empty pcDNA3 or a vector expressing RNPC1a, ΔRNP1, or ΔRNP2 for 24 h.

Figure 4.

RNPC1a interacts with eIF4E. (A,B) Cell lysates were prepared from HCT116 cells uninduced or induced to express RNPC1a for 18 h, treated with RNase A, and then immunoprecipitated with anti-HA (A) or anti-eIF4E (B) along with a control IgG. The immunocomplexes were examined with anti-eIF4E (A, top panel) or anti-HA (B, top panel). The blots were then stripped and reblotted with anti-HA (A, bottom panel) or anti-eIF4E (B, bottom panel). (C) HA-tagged RNPC1a physically interacts with Myc-tagged eIF4E proteins in vitro. HA-tagged RNPC1a or Myc-tagged eIF4E expression vector (1 μg) was used for coupled in vitro transcription/translation reactions as described in the Materials and Methods. Equal volumes of in vitro translated RNPC1 and eIF4E protein lysates were treated separately with RNaseA and then mixed, followed by immunoprecipitation with 1 μg of HA antibody or control IgG. The immunocomplexes were examined by Western blot analysis with anti-myc (top panel), and the blots were then stripped and reblotted with anti-HA (bottom panel). (D) GST pull-down assays were performed with recombinant GST-eIF4E and histidine-tagged RNPC1a. Equal amounts of GST or GST-fused eIF4E were incubated with his-tagged RNPC1a along with glutathione sepharose for 3 h. Complexes were then washed, followed by Western blot analysis using antibody against histidines (anti-omini) or GST. (E,F) GST pull-down assays were performed with recombinant GST-eIF4E and HA-tagged ΔRNP1 (E) or ΔRNP2 (F). The experiments were performed as in D, except that recombinant HA-tagged ΔRNP1 or ΔRNP2 protein was used. (G) GST pull-down assays were performed with recombinant GST-eIF4E and His-tagged RNPC1b. The experiments were performed as in D, except that recombinant His-tagged RNPC1b protein was used. (H,I) eIF4E does not interact with HA-tagged FR. The experiments were performed as in A and B, except that HCT116 cells were uninduced or induced to express HA-tagged FR.

Figure 5.

The p53 5′ UTR or a poly(U) element in the p53 3′ UTR is required for RNPC1 to inhibit p53 expression. (A) Schematic presentation of p53 mRNA and the location of probes. Poly(U) and AU-rich elements are indicated. (B) RNPC1 directly binds to the p53 5′ UTR. ³²P-labeled RNA probes were mixed with recombinant GST or HA-RNPC1a-GST or HA-RNPC1b-GST fusion protein. For competition assay, cold p21 probe was added to the reaction run in lanes 3 and 5, respectively. The bracket indicates RNA–protein complexes (RPC). (C) RNPC1 directly binds to the p53 3′ UTR. REMSA assay was performed as in B using the p53 3′ UTR as a probe. (D,E) RNPC1a directly binds to probe B but not probe A. REMSA assay was performed as in B with probes A and B. (F) The RNA-binding domain in RNPC1 is required for RNPC1a to bind p53 mRNA. REMSA assay was performed as in B by incubating ³²P-labeled probe B with GST, GST-HA-RNPC1a, GST-HA-ΔRNP1, or GST-HA-ΔRNP2. (G,H) The poly(U) element in the p53 3′ UTR is required for RNPC1a to bind p53 mRNA. REMSA assay was performed as in B with probes B and B1–B5. (I) The presence of the p53 5′ UTR or 3′ UTR is sufficient for RNPC1a to inhibit p53 expression. Various amounts of RNPC1a expression vector were transfected into H1299 cells along with a fixed amount of p53 expression vector that contains the coding region (ORF) alone or in combination with the 5′ UTR, the 3′ UTR, or both. The levels of p53, RNPC1a, and actin were analyzed by Western blot analysis. The fold change of p53 is shown below each lane.

Figure 6.

Loss of RNPC1 triggers p53-dependent senescence in primary MEFs. (A) Primary RNPC1^+/+, RNPC1^+/−, and RNPC1^−/− MEFs at passage 5 were treated with or without doxorubicin for 12 h, and the levels of RNPC1, p53, PAI-1, p130, p21, and GAPDH were analyzed by Western blot analysis. (B) Primary RNPC1^+/+, RNPC1^+/−, and RNPC1^−/− MEFs at passage 5 were treated with or without 50 ng/mL doxorubicin for 2 d, followed by SA-β-gal staining assay as described in the Materials and Methods. (C) Quantification of the percentage of SA-β-gal-positive cells as shown in B. (D) Primary wild-type, RNPC1^−/−, and RNPC1^−/−; p53^−/− MEFs at passage 5 were treated with or without doxorubicin for 12 h, and the levels of RNPC1, p53, PAI-1, p130, p21, and GAPDH were analyzed by Western blot analysis. (E) Primary wild-type, RNPC1^−/−, and RNPC1^−/− p53^−/− MEFs at passage 5 were treated with or without 50 ng/mL doxorubicin for 2 d, followed by SA-β-gal staining assay. (F) Quantification of the percentage of SA-β-gal-positive cells as shown in E.

Figure 7.

RNPC1 mRNA and protein are overexpressed in dog lymphomas. (A,C,E) The levels of RNPC1, p53, and actin in two normal dog lymph nodes and 28 dog lymphomas were measured by Western blot analysis. The levels of p53 and RNPC1a proteins in LN-3 (normal lymph node) were arbitrarily set at 1.0, and the fold change is shown below each lane. (B,D,F) The levels of RNPC1 and actin transcripts in normal and lymphoma tissues were measured by RT–PCR. The level of RNPC1a transcript in LN-3 (normal lymph node) was arbitrarily set at 1.0, and the fold change is shown below each lane. (G) A model for RNPC1 to regulate p53 translation.