Interplay between PTB and miR-1285 at the p53 3'UTR modulates the levels of p53 and its isoform Δ40p53α

Abstract

p53 and its translational isoform Δ40p53 are involved in many important cellular functions like cell cycle, cell proliferation, differentiation and metabolism. Expression of both the isoforms can be regulated at different steps. In this study, we explored the role of 3'UTR in regulating the expression of these two translational isoforms. We report that the trans acting factor, Polypyrimidine Tract Binding protein (PTB), also interacts specifically with 3'UTR of p53 mRNA and positively regulates expression of p53 isoforms. Our results suggest that there is interplay between miRNAs and PTB at the 3'UTR under normal and stress conditions like DNA damage. Interestingly, PTB showed some overlapping binding regions in the p53 3'UTR with miR-1285. In fact, knockdown of miR-1285 as well as expression of p53 3'UTR with mutated miR-1285 binding sites resulted in enhanced association of PTB with the 3'UTR, which provides mechanistic insights of this interplay. Taken together, the results provide a plausible molecular basis of how the interplay between miRNAs and the PTB protein at the 3'UTR can play pivotal role in fine tuning the expression of the two p53 isoforms.

Figure 1.

PTB binds specifically to p53 3′UTR. (A) α-³²P-labeled 3′UTR (Lanes 3,4) and α-³²P-labeled 5′UTR (lanes 5 and 6) were UV cross-linked with increasing concentration of PTB (250 ng, 500 ng), lane 2 is no protein (NP) and lane 1 represents the MW marker in kDa. (B) Competition UV crosslinking of PTB and α-³²P-labeled 3′UTR with 100- and 200-fold molar excess of either unlabeled p53 3′UTR RNA (lanes 4 and 5), unlabeled p53 5′UTR (lanes 6 and 7) or non-specific RNA (Nsp RNA) (lanes 8 and 9) showing the specificity of PTB and 3′UTR binding. Lane 1 represents the MW marker in kDa, lane 2 is no protein (NP) and lane 3 represents binding in the absence of competitor RNA (NC). (C) After UV cross-linking of α-³²P-labeled p53 3′UTR RNA with A549 cell S10 extract, immunoprecipitation was carried out using the respective antibodies, followed by incubation with protein G-Sepharose beads. The beads were either used after saturation with IgG isotype control antibody (lane 2) or anti-PTB antibody (lane 3). Lane 1, no protein (NP); lane 4, 25% of the UV-cross-linked S10 extract (input). Lane 5 is UV-crosslinked recombinant PTB (500 ng) protein. Numbers to the left represent relative mobilities of the molecular mass markers. (D) H1299 cells were transfected with Luc-p53 3′UTR construct. After 48 h, the cells were lysed with lysis buffer. RNP complexes were immunoprecipitated with anti- PTB antibody. p53 3′UTR was detected by RT-PCR analysis (n = 3).

Figure 2.

Partial silencing of PTB decreases p53 expression and target mRNA levels. (A) H1299 (p53) null cell line was co-transfected with luciferase (Firefly luciferase) constructs having only 5′UTR (5′UTR Luc), 3′UTR (Luc 3′UTR), or both 5′ and 3′UTR (5′UTR Luc 3′UTR), and either non specific si (si Nsp) or si PTB. (25 nM) Rluc (Renilla luciferase) was used as transfection control in all sets. Cells were harvested after 48 h in passive lysis buffer and processed for luciferase assay. The graph represents the normalized luciferase activity, i.e. Fluc/Rluc (F/R) (n = 3). (B) H1299 cells were co-transfected with p53 cDNA construct having 3′UTR (5′UTR + cDNA + 3′UTR) or no 3′UTR (5′UTR + cDNA), and either non specific si (si Nsp) or si PTB (50nM). No siRNA (No si) transfection is the control lane. After 48 h, the cells were lysed and processed for immunoblotting. Upper panel: p53 and Δ40p53; middle panel: PTB; lower panel: actin. (C) Western blot analysis of cell extracts from A549-cells transfected with either non specific si (si Nsp) or si PTB (100 nM), probed with CM1 after 48 h. Upper panel: p53 and Δ40p53; middle panel: PTB; lower panel: actin. (D) Quantitative PCR of p21 and 14–3-3σ mRNA levels normalized to actin in A549 cells transfected with either non specific si (si Nsp) or si PTB (100 nM) for 48 h (n = 4). (E) Quantitative PCR of p21 and 14–3-3σmRNA levels normalized to Actin in A549 cells transfected with either non specific si (si Nsp) or si PTB (100 nM) for 48 h and treated with doxorubicin (2 μM) for last 16 h (n = 5). (F) Western blot analysis of cell extracts from H1299 cells co-transfected with p53 cDNA construct having 3′UTR (5′UTR + cDNA + 3′UTR) and either non specific si (si Nsp) or si PTB (100 nM), probed with CM1 after 48 h post transfection. (G) Quantitative PCR of p21 and 14–3-3σ mRNA levels normalized to GAPDH in H1299 cells co-transfected with p53 cDNA construct having 3′UTR (5′UTR + cDNA + 3′UTR) and either non specific si (si Nsp) or si PTB (100 nM) for 48 h (n = 3).

Figure 3.

PTB binding regions in p53 3′UTR. (A) Mfold-predicted structure of complete p53 3′UTR. (B) Different regions of p53 3′UTR used in panel C are shown separately. (C) Competition UV crosslinking of α-³²P-labeled 3′UTR and unlabeled cold RNA of different regions of 3′UTR with PTB. Lane 1: no protein (NP), lane 2: no competition (NC), lane 3 and 4: 100- and 200-fold molar excess of either unlabeled p53 3′UTR region I RNA, lanes 5 and 6: 100- and 200-fold molar excess of either unlabeled p53 3′UTR region II RNA, Lanes 7 and 8: 100- and 200-fold molar excess of either unlabeled p53 3′UTR region III RNA, lanes 9 and 10: 100- and 200-fold molar excess of either unlabeled p53 3′UTR region IV RNA, lanes 11 and 12: 100- and 200-fold molar excess of either unlabeled p53 3′UTRRNA, lanes 13 and 14: non-specific RNA (Nsp RNA).

Figure 4.

Interplay between the ITAFs and miRNAs at 3′UTR. (A) H1299 cells were co-transfected with Luc-3′UTR and either si PTB/non specific siRNA (50 nM). After 48 h, the cells were lysed with lysis buffer. RNP complexes were immunoprecipitated with anti- Ago-2/IgG antibody. p53 3′UTR was detected by RT-PCR analysis. The values for RNA immunoprecipitated with antiAgo-2 antibody in different conditions (si Nsp and si PTB) were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody and then the fold change was calculated using the value for 3′UTR pulled down with anti-Ago-2 upon si Nsp transfection as the basal value (n = 4). (B) H1299 cells were transfected individually with the reporter constructs Fluc (labeled as Luc), Fluc-3′UTR (labeled as Luc3′UTR). These cells were treated with doxorubicin (2 μM) for 16 h. After 24 h, the cells were harvested and checked for luciferase activity (n = 5). (C) H1299 cells were transfected with plasmid expressing luciferase-p53 3′UTR (Luc 3′UTR). The cells were treated with doxorubicin (2μM) for 16 hours. After 24 hours the cells were harvested. RNP complexes were immunoprecipitated with anti-Ago-2 antibody. p53 3′UTR was detected by RT-PCR analysis. The values for RNA immunoprecipitated with anti Ago-2 antibody in different conditions (+/– DOX) were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody and then the fold change was calculated using the value for 3′UTR pulled down with anti-Ago-2, without doxorubicin treatment as the basal value (n = 5). (D) RNP complexes were immunoprecipitated with anti- Ago-2 antibody and endogenous p53 3′UTR was detected by RT-PCR analysis from A549 cells in presence and absence of doxorubicin (2 μM) for 16 h. The values for RNA immunoprecipitated with anti Ago-2 antibody in different conditions (+/– DOX) were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody and then the fold change was calculated as in Figure 4C (n = 5). (E) H1299 cells were transfected with plasmid expressing luciferase-p53 3′UTR (Luc 3′UTR) or luciferase (Fluc) alone. The cells were treated with doxorubicin (2 μM) for 16 h. After 24 h, the cells were harvested. RNP complexes were immunoprecipitated with anti-PTB antibody. Luciferase mRNA was detected by RT-PCR analysis. The values for RNA immunoprecipitated with anti PTB antibody under different conditions (+/– DOX) were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody, and then the fold change was calculated using the value for 3′UTR pulled down with anti-PTB, without doxorubicin treatment as the basal value (n = 3). (F) RNP complexes were immunoprecipitated with anti-PTB antibody and endogenous p53 3′UTR was detected by RT-PCR analysis from A549 cells in the presence and absence of doxorubicin (2 μM) for 16 h. The values for RNA immunoprecipitated with anti PTB antibody in different conditions (+/– DOX) were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody and then the fold change was calculated as in D (n = 4).

Figure 5.

Effect of knockdown of miRNAs on the expression of p53 isoforms and miRNAs interplay with PTB. (A) H1299 cells were co-transfected with Luc-3′UTR and anti-miRs for miR-30d/miR-1285/miR-504/miR-181 (10 nM) and Rluc (Renilla luciferase) as transfection control. After 48 h, the cells were lysed and processed for luciferase activity. The graph represents the normalized luciferase activity, i.e. Fluc/Rluc (F/R) (n = 3). (B) H1299 cells were co-transfected with p53 cDNA construct having 3′UTR (5′UTR + cDNA + 3′UTR) or no 3′UTR (5′UTR + cDNA) and anti-miR-30d/anti-miR-1285 in two concentrations (10 nM and 20 nM). After 48 h, the cells were lysed and processed for immunoblotting. (C) Western blot analysis of cell extracts from A549-cells transfected with anti-miRs for miR-1285/miR-30d (30 nM), probed with CM1 after 48 h. Upper panel: p53 and Δ40p53; lower panel: actin. (D) Quantitative PCR of p21 and 14–3-3σ mRNA levels normalized to Actin in A549 cells transfected with anti-miRs for miR-1285/ miR-30d (30 nM) for 48 h (n = 3). (E) H1299 cells were co-transfected with Luc-3′UTR and anti-miRs for miR-30d/miR-1285 (20 nM). After 48 h, the cells were lysed and RNP complexes were immunoprecipitated with anti-PTB antibody. RT PCR was performed for checking the levels of 3′UTR associated with PTB. The values for RNA immunoprecipitated with anti PTB antibody in different conditions (no anti-miR, anti-miR-30d and anti-miR-1285) were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody and then the fold change was calculated using the value for 3′UTR pulled down with anti-PTB, without anti-miR treatment as the basal value (n = 5). (F) H1299 cells were transfected with wild type Fluc -3′UTR (labeled as WT), Fluc -3′UTR mut miR-1285 M1 (3′UTR mutated for miR-1285 one binding site, labeled as M1), Fluc -3′UTR mut miR-1285 M1+M2 (3′UTR mutated for miR-1285 both binding sites, labeled as M1 + M2) and GFP expression from a GFP plasmid vector was used as transfection efficiency control. After 24 h, the cells were lysed and processed for luciferase activity. The graph shows normalized Fluc activity of different constructs (n = 3). (G) H1299 cells were transfected with either wild type Fluc -3′UTR (labeled as WT) or Fluc -3′UTR mut miR-1285 M1+M2 (3′UTR mutated for miR-1285 both binding sites, labeled as M1 + M2). After 24 h, the cells were lysed and RNP complexes were immunoprecipitated with anti-PTB antibody. RT PCR was performed for checking the levels of 3′UTR associated with PTB. The values for RNA (WT and M1 + M2) immunoprecipitated with anti PTB antibody were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody, and then the fold change was calculated using the value for 3′UTR pulled down with anti-PTB from WT-3′UTR transfected cells as the basal value (n = 6). (H) H1299 cells were transfected with either wild type Fluc -3′UTR (labeled as WT) or Fluc -3′UTR mut miR-1285 M1+M2 (3′UTR mutated for miR-1285 both binding sites, labeled as M1 + M2). After 24 h, the cells were lysed and RNP complexes were immunoprecipitated with anti-Ago-2 antibody. q-RT PCR was performed for checking the levels of 3′UTR associated with Ago-2. The values for RNA (WT and M1+M2) immunoprecipitated with anti Ago-2 antibody were first normalized with its respective values for RNA immunoprecipitated with IgG control antibody and then the fold change was calculated using the value for 3′UTR pulled down with anti-Ago-2 from WT-3′UTR transfected cells as the basal value (n = 6). (I) Toe-printing analysis to map the contact points of PTB on p53 3′UTR. p53 3′UTR RNA was incubated in absence (lane 7) and presence (lanes 5 and 6) of increasing concentrations of purified recombinant PTB (200 ng, 400 ng). The RNA in the ribonucleoprotein complexes were reverse transcribed using 3R3 reverse primer (p53 3′UTR region 3 reverse primer) and the resulting cDNAs were resolved in 8% acrylamide–8 M urea PAGE in parallel with a sequencing reaction. The cDNA products terminated at the sites due to protein binding is marked by the arrows. The toe-prints at positions 61 (749T), 55 (755T), 54 (756T), 52 (758T) in the 3′ UTR are indicated. (J) Schematic representation of the toe-prints of PTB protein on putative predicted secondary structure of p53 3′UTR R-III (third region) RNA (nucleotides 759–809 in 3′UTR) generated by MFOLD. Nucleotides 748 to 758 bind PTB and are pointed out in red. The curly bracket region shows the miR-1285 binding site (seed sequence). (K) α-³²P-labeled 3′UTR (wild type: WT, mutant: 2a = TT to GG 755–756, 3a = TT to GG 748–749) were UV crosslinked with increasing concentration of PTB (50 ng, 100 ng), lane 1 is no protein (NP). (L) H1299 cells were transfected with wild type Luc -3′UTR (labeled as WT), Luc -3′UTR Mut2a and Luc -3′UTR Mut3a (Mutant: 2a = TT to GG 755–756, 3a = TT to GG 748–749) and Rluc was used as transfection control. After 24 h, the cells were lysed and processed for luciferase activity. The graph shows the normalized Fluc activity (F/R) of different constructs (n = 6).

Figure 6.

SNVs in the p53 3′UTR can affect 3′UTR function. (A) A549 cells were transfected with luciferase constructs having either wild type 3′UTR or 3′UTR with the SNVs (SNV 287, SNV 737, SNV 93 and SNV 806). After 12 h and 24 h, the cells were lysed and processed for luciferase activity (n = 6). (B) H1299 cells were transfected with luciferase constructs having either wild type 3′UTR or 3′UTR with the SNVs (SNV 287, SNV 737, SNV 93 and SNV 806). After 12 h and 24 h, the cells were lysed and processed for luciferase activity (n = 6). (C) Cytoplamic-S10 extract from A549 cells at three different concentrations) was incubated with α-³²P UTP labeled p53 3′UTR WT and p53 3′UTR having the SNV 806 (C-T) RNA for UV cross-linking. Protein bands in the box show that PTB binding in both S10 and recombinant protein is more with p53 3′UTR having the SNV 806 (C-T) as compared to WT 3′UTR.

Figure 7.

Model depicting the interplay between the proteins and miRNAs. (A–C) Model showing the interplay between the proteins (PTB) and miRNAs (miR-1285) at the p53 3′UTR resulting in the fine tuning of the expression of p53 isoforms.