Structure of tumor suppressor p53 and its intrinsically disordered N-terminal transactivation domain

TL;DR

Using NMR residual dipolar couplings (RDCs) and small-angle x-ray scattering (SAXS), the authors characterized the intrinsically disordered N-terminal domain of p53 and modeled its ensemble in isolated, full-length, DNA-bound, and Taz2-bound states. They found a stiff proline-rich region that projects the transactivation domain outward and a transient helical element in the TAD.

Key findings

• Isolated p53(1–93) is disordered with nascent structure: residues 22–25 adopt a helical conformation populated at ≈30%, and the proline-rich region (PRR, residues 64–92) shows elevated polyproline II (PPII) propensity, with optimal RDC-derived populations of 79 ± 2% for proline and 50 ± 2% for preproline residues, close to accelerated MD predictions of 78 ± 3% and 46 ± 4%.
• The PRR is stiffer than the TAD and projects the TAD away from the core domain; SAXS data for p53(1–93) were reproduced by the RDC-derived ensemble and were not consistent with a collapsed TAD.
• RDCs from the TAD were measured in isolated p53(1–93), full-length tetrameric p53, and p53–DNA; in DNA-bound flp53 the TAD couplings were smaller and PRR couplings larger, whereas DNA-free RDCs were similar to isolated p53(1–93) except for smaller PRR values.
• SAXS showed the flp53–Taz2 complex has four Taz2 domains per tetramer and is an extended cross-shaped structure fitted with χ = 1.19; a ternary flp53–Taz2–DNA model fitted with χ = 2.0, with Taz2-bound TADs projecting outward from the core and core domains enveloping the response element.

Methods

• RDC NMR experiments used strained polyacrylamide gel and pf1 bacteriophage alignment at 20°C on isolated p53(1–93), a full-length p53 quadruple mutant, p53(1–393)–DNA, and synthetic 15N-labeled peptides comprising residues 61–93; measured couplings included 1DNHN, 2DC′HN, 1DC′Cα, 1DCαHα, and 3JHαHN.
• SAXS data were collected at the X33 beamline at EMBL/DESY, Hamburg; p53(1–93) samples ranged from 2 to 8.6 mg/ml; data were processed with PRIMUS and modeled with ATSAS programs CRYSOL, SASREF, and BUNCH, including rigid-body refinement using the Taz2 structure PDB 1F81.
• Conformational ensembles were generated with flexible-meccano sampling of coil-library propensities and accelerated MD simulations (AMBER8) for PRR residues 70–82; 50,000 tetramer structures were built with four independent N-terminal chains (200,000 chains total), and RDCs were predicted and averaged over the ensemble.

Limitations

• Stated: modeling the N terminus in the context of DNA-free flp53, which contains several rigid bodies in a highly dynamic state, was beyond the scope of the study; the absolute level of alignment is difficult to predict from simulation; and SAXS reproduction deviated at q < 0.04, where experimental values were higher than predicted.
• Evident: conclusions about the disordered N-terminal domain rely on ensemble and rigid-body models fitted to solution RDC/SAXS data rather than single high-resolution structures; the models are described as probably representative of dynamic ensembles, so conformational heterogeneity may not be fully resolved.