p28, a first in class peptide inhibitor of cop1 binding to p53

Abstract

Background: A 28 amino-acid (aa) cell-penetrating peptide (p28) derived from azurin, a redox protein secreted from the opportunistic pathogen Pseudomonas aeruginosa, produces a post-translational increase in p53 in cancer cells by inhibiting its ubiquitination.

Methods: In silico computational simulations were used to predict motifs within the p53 DNA-binding domain (DBD) as potential sites for p28 binding. In vitro direct and competitive pull-down studies as well as western blot and RT-PCR analyses were used to validate predictions.

Results: The L1 loop (aa 112-124), a region within the S7-S8 loop (aa 214-236) and T140, P142, Q144, W146, R282 and L289 of the p53DBD were identified as potential sites for p28 binding. p28 decreased the level of the E3 ligase COP1 >80%, in p53wt and p53mut cells with no decrease in COP1 in p53dom/neg or p53null cells. Brief increases in the expression of the E3 ligases, TOPORS, Pirh2 and HDM2 (human double minute 2) in p53wt and p53mut cells were in response to sustained increases in p53.

Conclusion: These data identify the specific motifs within the DBD of p53 that bind p28 and suggest that p28 inhibition of COP1 binding results in the sustained, post-translational increase in p53 levels and subsequent inhibition of cancer cell growth independent of an HDM2 pathway.

Figure 1

Docking model of p28 and p53. (A) The p28–p53 complex structure obtained from the best docking model. Blue: overall ribbon diagram of the p53DBD, green: L₁ and S₇–S₈ loops. (B) Ribbon diagram of the p53DBD (yellow) superimposed on the electrostatic potential plot. p28 (green) binds to hydrophobic pocket. (C) Relationship of p28 (green) and L₁ (blue) and S₇–S₈ (yellow) loops of the p53DBD. (D) Map of E3 ligase binding regions superimposed on the full-length sequence of p53. p28-binding residues indicated in red. Binding regions of HDM2, COP1 and Pirh2 indicated in black, green and blue bars, respectively. COP1 binds to p53 in regions separate from HDM2. p28 and COP1 share a binding motif within aa 92–160 of the p53DBD.

Figure 2

p28-binding regions on p53. (A) Purified various fragments of p53 and GST alone were used for a GST pull-down assay. Each sample was analysed by immunoblotting with an anti-p28 antibody. (B) Competitive pull-down assay. Immobilised GST-p53_1–393 and GST alone on glutathione-Sepharose 4B beads were incubated in absence (−) or presence of p28 (+: 10, ++: 100 mole excess), followed by adding MCF-7 lysates containing COP1. Samples were separated by SDS-PAGE and immunoblotted with an anti-COP1 antibody. Lysate: whole-cell lysates of MCF-7 used for the assay showed the stable expression of COP1. Numbers below COP1 bands were relative percentage to the level of COP1 bound to p53 in absence of p28. (C) COP1 binding to various fragments of p53 in absence or presence of p28 was analysed using a similar competitive pull-down assay. Immobilised various fragments of p53 and GST alone were treated with or without p28 (100 mole excess), followed by incubation with MCF-7 lysates containing COP1.

Figure 3

The effect of p28 on cell and xenograft growth. Human melanoma UISO-Mel-29, -23 and -6 (A) and breast cancer MCF-7, MDD2, MDA-MB-231 and normal MCF-10A cells (B) were exposed to p28, DTIC, paclitaxel or a similar volume of media without peptide for 72 h and cells counted. Values represent the mean±s.d. of three replicates. Reduction in Mel-23, 29 xenograft growth (C). Reduction of MCF-7, MDA-MB-231 xenograft growth (D). *P<0.05.

Figure 4

Stabilisation of p53 by E3 ligases on p53^wt,mut breast cancer cells. (A) MCF-7, MDD2, MDA-MB-231, and MCF-10A cells were treated with p28 at 50 μmol l⁻¹ for 24–72 h. Whole-cell lysates were subjected to western blot analysis. (B) The expression of each gene was determined by RT-PCR. The numbers indicated below each band represent the relative to control (control expressed as 100%).

Figure 5

Stabilisation of p53 by E3 ligases on p53^wt,mut melanoma cells. (A) Mel-29, Mel-23 and Mel-6 cells were treated with p28 at 50 μmol l⁻¹ for 24–72 h. (B) The expression of each gene was determined by RT-PCR. The numbers indicated below each band represent the relative to control (control expressed as 100%).

Figure 6

The effect of p28 on c-Jun. The protein level of c-Jun was determined in MCF-7, MDD2, MDA-MB-231 and MCF-10A cells (A), and Mel-29, Mel-23 and Mel-6 (C) treated with p28 at 50 μmol l⁻¹ for 24–72 h. The transcription level of c-Jun was determined by RT-PCR in breast cell lines (B) and melanoma cells (D). The numbers indicated below each band represent the relative to control (control expressed as 100%).