Probing potential binding modes of the p53 tetramer to DNA based on the symmetries encoded in p53 response elements

Abstract

Symmetries in the p53 response-element (p53RE) encode binding modes for p53 tetramer to recognize DNA. We investigated the molecular mechanisms and biological implications of the possible binding modes. The probabilities evaluated with molecular dynamics simulations and DNA sequence analyses were found to be correlated, indicating that p53 tetramer models studied here are able to read DNA sequence information. The traditionally believed mode with four p53 monomers binding at all four DNA quarter-sites does not cause linear DNA to bend. Alternatively, p53 tetramer can use only two monomers to recognize DNA sequence and induce DNA bending. With an arrangement of dimer of AB dimer observed in p53 trimer-DNA complex crystal, p53 can recognize supercoiled DNA sequence-specifically by binding to quarter-sites one and four (H14 mode) and recognize Holliday junction geometry-specifically. Examining R273H mutation and p53-DNA interactions, we found that at least three R273H monomers are needed to disable the p53 tetramer, consistent with experiments. But just one R273H monomer may greatly shift the binding mode probabilities. Our work suggests that p53 needs balanced binding modes to maintain genome stability. Inverse repeat p53REs favor the H14 mode and direct repeat p53REs may have high possibilities of other modes.

Figure 1.

Definition of quarter-site coupling (A) and five corresponding binding modes. (B) Q1234 is the mode of fully occupied quarter-sites with half-site palindrome. (C) T13 mode. (D) T24 mode. (E) The H14 mode uses quarter-sites 1 and 4 to tightly bind p53, and (F) H23 where quarter-sites 2 and 3 are fully occupied. Starting models assembled from existing crystal structures for four binding modes (Q1234, T13, T24 and H14) are illustrated. The initial structures are subjected to refinements using molecular dynamics simulations.

Figure 2.

Change of linearity of the DNA double helix during molecular dynamics simulations of the tetramer bound to the DNA in the H14 mode with the p53 interacting with four p53REs (A), p21-5′ binding site; (B) GADD45; (C) pDINP1; and (D) p53AIP1. A potential mechanism in these simulations for cooperative p53–DNA interaction and DNA bending is illustrated in box E.

Figure 3.

Structural details of the potential H14 binding mode, in which two swapped copies of p53 dimer bind DNA symmetrically in respect to full-site palindrome. p53-tetramer–Puma BS2 complex is used here. Chains binding DNA specifically at quarter-sites one and four are the B chains in the p53-trimer–DNA complex. B1 and B2 are used to term their position in two dimers, respectively. The A chains assist in the DNA recognition and provide tetramer stability. A1 and A2 are used to name the chain position. (A) A view along the DNA chain. (B) A view perpendicular to the DNA chain, in which the p53 tetramer is inside the DNA loop. (C) A view from the top of the complex illustrating the arrangement of the core domain tetramer. (D) Atomic details of the salt bridges stabilizing the p53 tetramer.

Figure 4.

Trajectories of prolonged simulations of PUMA BS2 binds p53 in Q1234 and H14 modes. (A) Total energies of simulated systems. (B) RMSDs of backbone atoms of p53 tetramer in Q1234 and H14 modes, respective to the structures averaged from first 10 ns simulations. (C) Two-dimensional RMSD map for the H14 mode. 42 conformations are obtained by average conformations during 1 ns windows in the 42 ns trajectory. Then the RMSD between all pair of conformers are calculated. The red dots indicated the RMSD positions among three conformations (12 ns, 22 ns and 37 ns). (D) The superimposed structures for the three conformations are illustrated: 12 ns, redline; 22 ns greenline; 37 ns, blueline.

Figure 5.

Models of the p53–Holliday junction interactions. The model presents tetrameric arrangements similar to the DNA sequence-specific p53 core domain (Figures 3 and 4). (I) shows p53 core domain tetramer interacting with the Holliday junction; (II) trajectories of MD simulation of Holliday junctions binding with two separated p53 dimers (Supplementary Figure 5B) and p53 tetramer (Figure 5I). (III) Illustrates the p53 tetramer binding the DNA supercoil and the Holliday junction using the same arrangement.

Figure 6.

(A) The correlation of the binding mode probability derived from statistics of DNA sequences of p53REs (Y-axis) with the probability derived from molecular simulations (X-axis). The detailed values are provided in Table 1. The points are colored according to DNA sequences (Red: p21-5′, Blue: Gadd45, Green: pDINP1, Violet: p53AIP1, Pink: Puma BS2 and shaped according to binding modes (Triangle: Q1234, Square: T24, Diamond: T13, Circle: H14). (B). Change of binding mode probability with increasing of R273H mutant in p53-tetramer–Puma BS2 complex. p53 is able to have attractive interaction with DNA when two monomers in the tetramer are wild-type p53 (Table 2).

Figure 7.

Upper panel: EM image of p53 tetramer in p53–DNA complex in solution, reference (48). (Reproduced with permission). The see-through channel in the EM map indicates the possible DNA position. Lower panel: fit of p53–DNA complex in H14 binding mode to the EM image in the upper panel. The tetramerization domain was added to the Puma BS2–p53 core domain tetramer in Figure 3.