Sequence Properties of An Intramolecular Interaction That Inhibits p53 DNA Binding

Abstract

An intramolecular interaction between the p53 transactivation and DNA binding domains inhibits DNA binding. To study this autoinhibition, we used a fragment of p53, referred to as ND WT, containing the N-terminal transactivation domains (TAD1 and TAD2), a proline rich region (PRR), and the DNA binding domain (DBD). We mutated acidic, nonpolar, and aromatic amino acids in TAD2 to disrupt the interaction with DBD and measured the effects on DNA binding affinity at different ionic strengths using fluorescence anisotropy. We observed a large increase in DNA binding affinity for the mutants consistent with reduced autoinhibition. The ΔΔG between DBD and ND WT for binding a consensus DNA sequence is -3.0 kcal/mol at physiological ionic strength. ΔΔG increased to -1.03 kcal/mol when acidic residues in TAD2 were changed to alanine (ND DE) and to -1.13 kcal/mol when all the nonpolar residues, including W53/F54, were changed to alanine (ND NP). These results indicate there is some cooperation between acidic, nonpolar, and aromatic residues from TAD2 to inhibit DNA binding. The dependence of DNA binding affinity on ionic strength was used to predict excess counterion release for binding both consensus and scrambled DNA sequences, which was smaller for ND WT and ND NP with consensus DNA and smaller for scrambled DNA overall. Using size exclusion chromatography, we show that the ND mutants have similar Stokes radii to ND WT suggesting the mutants disrupt autoinhibition without changing the global structure.

Keywords: DNA binding; counterion condensation theory; fluorescence anisotropy; hydrodynamic radius; intramolecular interaction; intrinsically disordered proteins; salt-dependent binding affinity; tumor suppressor p53; van’t Hoff.

Figure 1

p53’s disordered TAD2 interacts with DBD. (a) A domain map shows p53’s domains. (b) IUPRED plot of full length p53 WT predicts regions of disorder based on sequence. The red box defines the region containing TAD2. (c) Inset from red box in (b) of IUPRED plot of a region containing TAD2 compares the disorder prediction of the wild type TAD2 and three mutants, where residues above the 0.5 line are predicted to be disordered. (d) Agadir prediction of helical propensity of the TAD2 region using wild type TAD2 and three mutants. (e) TAD2 sequences of the WT and mutants used in this study; red boxes indicate negatively charged residues, green boxes indicate polar residues, and gold boxes indicate nonpolar residues. (f) TAD2 interacts with DBD, inhibiting DNA binding by a combination a charge-based and specific interactions.

Figure 2

DBD binds DNA across IS. Fluorescence anisotropy plots show the change in signal from a fluorescently tagged DNA fragment as protein is added: an increase in the concentration of p53 needed to achieve saturation when DNA concentration is kept stable as buffer salt concentration increases. (a) fluorescence anisotropy plots of DBD bound to consensus DNA at 125–225 mM IS; (b) fluorescence anisotropy plots of DBD bound to scrambled DNA at 125–225 mM IS.

Figure 3

Binding of DBD and ND fragments to consensus and scrambled DNA at physiological IS (145 mM). (a) Fluorescence anisotropy plots of p53 constructs binding consensus DNA, where is DBD, is ND WT, is ND DE, is ND NP, is ND QS, (b) p53 constructs binding scrambled DNA, where is DBD, is ND WT, is ND DE, DNA, is ND NP, is ND QS, (c) ΔG of all fragments with consensus and scrambled DNA. Each data set represents three titrations.

Figure 4

Binding specificity of DBD, ND WT, and ND mutants. For each p53 fragment, ΔΔG = ΔG_consensus − ΔG_scrambled at a given IS indicates binding specificity.

Figure 5

Salt-dependent binding affinity of DBD and ND WT. Plot of log (K_A) vs. log [Salt] from 125–225 mM IS of (a) DBD and ND WT binding to consensus DNA where is DBD, is ND WT, (b) DBD and ND WT binding to scrambled DNA where is DBD, is ND WT. R² values for all fit lines are between 0.96 and 0.99.

Figure 6

Salt-dependent binding affinity of ND mutants. Plot of log (K_A) vs. log[Salt] from 125–225 mM IS of (a) ND mutants binding consensus DNA, where is ND DE, is ND NP, is ND QS (b) ND mutants binding scrambled DNA, where is ND DE, is ND QS. Inset shows ND NP binding scrambled DNA, . R² values for all fit lines are between 0.96 and 0.99.

Figure 7

Salt-dependent and salt-independent components of Gibbs free energy at 145 mM IS from Record’s model. Free energy is apportioned into categories by assuming complete inhibition of salt-dependent components at 1M NaCl so the remainder of free energy is salt-independent. Slopes of double log plot slopes are used to estimate binding affinity at 1M NaCl, where is the salt-dependent component and is the salt-independent component for consensus DNA and is the salt-dependent component and is the salt-independent component for scrambled DNA.

Figure 8

Size exclusion chromatography is used to compare p53 constructs. Elution profiles of p53 constructs where lower elution volume indicates a larger hydrodynamic radius: DBD, ND DE, ND NP, ND QS, ND WT.