Interaction between p53 N terminus and core domain regulates specific and nonspecific DNA binding

Abstract

The p53 tumor suppressor is a sequence-specific DNA binding protein that activates gene transcription to regulate cell survival and proliferation. Dynamic control of p53 degradation and DNA binding in response to stress signals are critical for tumor suppression. The p53 N terminus (NT) contains two transactivation domains (TAD1 and TAD2), a proline-rich region (PRR), and multiple phosphorylation sites. Previous work revealed the p53 NT reduced DNA binding in vitro. Here, we show that TAD2 and the PRR inhibit DNA binding by directly interacting with the sequence-specific DNA binding domain (DBD). NMR spectroscopy revealed that TAD2 and the PRR interact with the DBD at or near the DNA binding surface, possibly acting as a nucleic acid mimetic to competitively block DNA binding. In vitro and in vivo DNA binding analyses showed that the NT reduced p53 DNA binding affinity but improved the ability of p53 to distinguish between specific and nonspecific sequences. MDMX inhibits p53 binding to specific target promoters but stimulates binding to nonspecific chromatin sites. The results suggest that the p53 NT regulates the affinity and specificity of DNA binding by the DBD. The p53 NT-interacting proteins and posttranslational modifications may regulate DNA binding, partly by modulating the NT-DBD interaction.

Keywords: DNA binding; NMR; intramolecular; p53; specificity.

Fig. 1.

Detection of p53 domain binding. (A) Design of p53c1 and p53c2 constructs with Precision (PreS) cleavage sites and epitope tags (FLAG, HA, and Myc). (B) Diagram of proteolytic fragment release assay. (C and D) p53c2 expressed in H1299 was immobilized on FLAG or Myc beads and cleaved by PreScission for 30 min. The beads and supernatant were analyzed for the DBD fragment by HA Western blot (WB). Pairwise quantitation is shown below the blots. IP, immunoprecipitation. (E) p53c1 was immobilized on FLAG beads, cleaved by PreScission, and analyzed for DBD fragment release by WB using an antibody for full-length p53 (FL393). Pairwise quantitation is shown below the blots.

Fig. 2.

Analysis of p53 domain binding. (A) Diagram of WT, ALAL, and KEEK p53c1 constructs. PreS, Precision cleavage sites. (B) WT (Wt), ALAL, and KEEK p53c1 expressed in H1299 were immobilized on M2 beads, cleaved by protease, and analyzed for DBD fragment release by Western blot using the p53 antibody FL393. (C) Lysate of H1299 transfected with FLAG-p53-90–393 was incubated with beads loaded with GST or GST-p53-1–90. Pull-down of FLAG-p53-90–393 was detected by FLAG Western blot. (D) Lysate of H1299 transfected with FLAG-p53-90–393 was incubated with glutathione beads loaded with GST, GST-p53-1–90, GST-p53-1–40 (TAD1), GST-p53-35–65 (TAD2), and GST-p53-65–90 (PRR). Pull-down of FLAG-p53-90–393 was detected by FLAG Western blot.

Fig. 3.

p53 NT inhibits DNA binding in vitro. (A) Diagram of N-terminal–truncated p53c1 mutants with C-terminal FLAG tag. (B) FLAG-tagged p53c1 mutants were immunopurified from transfected H1299 cells, cleaved with protease, and incubated with beads loaded with biotinylated oligonucleotide containing the p53 binding site. The captured p53 was detected by Western blot using FL393. The level of p53 input was confirmed by Western blot. (C and D) Purified p53c1 mutants were cleaved with protease and incubated with immobilized biotinylated oligonucleotide in the presence of antibodies. The captured p53 was detected by Western blot using FL393. Pairwise quantitation is shown below the blots. (E) Purified FLAG-p53-90–393 was incubated with immobilized biotinylated oligonucleotide in the presence of NT peptides (200 μM). The captured p53 was detected by FLAG Western blot. Quantitation is shown below the blots.

Fig. 4.

p53 NT promotes dissociation from DNA. (A) Diagram of the in vitro luciferase fragment complementation assay for detecting p53 dissociation from DNA. (B) p53-Cluc was incubated with ZF-Nluc and ZPBS oligonucleotide for 30 min to assemble active complexes. Competitor oligonucleotide containing the p53 binding site was added, and luciferase activity was monitored for an additional 75 min. (C) Diagram of p53-Cluc fusion constructs with and without the protease cleavage site. (D) p53-Cluc and 90–393-Cluc were preincubated with ZF-Nluc and ZPBS oligonucleotide to assemble active complexes. Competitor oligonucleotide was added (0 min), and luciferase activity was monitored for 45 min. (E) Cleavable p53c1-Cluc and noncleavable p53-Cluc were preincubated with ZF-Nluc and ZPBS oligonucleotide in the presence and absence of protease to assemble active complexes. Competitor oligonucleotide was added (0 min), and luciferase activity was monitored for 45 min.

Fig. 5.

p53 NT inhibits DNA binding in vivo. (A) DNA binding affinity of the p53 ND fragment (1–312) and DBD (94–312) to the consensus DNA binding site and scrambled control DNA as determined by ITC. (B) p53 and N-terminal deletion mutants were expressed in H1299 cells using a lentiviral vector and induced to comparable levels using doxycycline. (C) p53 mutant expression was induced by doxycycline for 24 h. p53 binding to the p21 promoter was determined by ChIP. (D) WT p53 and the 90–393 mutant were induced in H1299 by doxycycline for 24 h. p53 occupancy at specific target promoters (average of p21, MDM2, PUMA, and Fas1) and nonspecific repetitive sequences (average of LINE1, Alu, and α-satellite) was determined by ChIP. The occupancy by WT p53 was arbitrarily set as onefold.

Fig. 6.

MDMX regulates p53 domain interaction and DNA binding. (A) p53-Cluc was preincubated with ZF-Nluc and ZPBS oligonucleotide to assemble active complexes in the presence of GST, GST-RPA-1–121, and GST-MDMX-1–120. Competitor oligonucleotide was added (0 min), and luciferase activity was detected at indicated time points. (B) MDMX disrupts p53 NT–DBD interaction. p53c1 was coexpressed with MDMX-1–200 and MDMX-1–490 in H1299 cells. p53c1 (and the associated MDMX) was immobilized on M2 beads, cleaved with protease for 40 min, and analyzed for NT–DBD interaction by Western blot detection of a released 91–393 fragment. Sup, supernatant. (C) U2OS cells stably infected with lentivirus expressing doxycycline-inducible MDMX were induced and treated with γ-radiation for 4 h. Expression of p53 pathway markers was determined by Western blot. IR, ionizing radiation. (D) U2OS cells expressing doxycycline-inducible MDMX were treated with doxycycline for 24 h and irradiated with 10 Gy IR. p53 binding to specific promoters (average of p21, MDM2, Fas1, and PUMA) and repetitive sequences (average of LINE1, Alu, and α-satellite) was analyzed by ChIP 4 h after irradiation. The occupancy by p53 in untreated cells was arbitrarily set as onefold. Con, control. *P < 0.05.

Fig. 7.

Intramolecular interaction between TAD2 and the DBD. (A) Overlay of ¹H-¹⁵N HSQC spectra for the apo (blue) and DNA-bound (red) ND. (B) Expanded view showing most of the disordered TAD1/2 resonances. (C) TAD2 and PRR residues with the largest chemical shift changes. (D) Plot of the average amide ¹H and ¹⁵N chemical shift changes for TAD1/2 and PRR residues in the apo vs. DNA-bound ND. The gray line marks the digital resolution of the HSQC experiment. (E) Intensity ratios for TAD1/2 and PRR residues calculated as DNA-bound divided by the apo ND. The gray line shows the expected intensity ratio if there is no interaction. (F) Transverse R2s of TAD1/2 and PRR residues in the ND (blue) and the NT (1–89, black).

Fig. 8.

TAD1/2 induces chemical shifts in the DBD. (A) Overlay of ¹H-¹⁵N HSQC spectra for the apo DBD (green) and ND (blue). (B) Plot of the average amide ¹H and ¹⁵N chemical shift changes for apo DBD and ND residues. Residues that contact DNA are indicated with red triangles, and red arrows show the positions of hotspot mutations. (C) Residues with the largest chemical shifts mapped onto the DBD tetramer (gray) bound to DNA (transparent cyan) (Protein Data Bank ID code 4hje).