Mechanism of rescue of common p53 cancer mutations by second-site suppressor mutations

Abstract

The core domain of p53 is extremely susceptible to mutations that lead to loss of function. We analysed the stability and DNA-binding activity of such mutants to understand the mechanism of second-site suppressor mutations. Double-mutant cycles show that N239Y and N268D act as 'global stability' suppressors by increasing the stability of the cancer mutants G245S and V143A-the free energy changes are additive. Conversely, the suppressor H168R is specific for the R249S mutation: despite destabilizing wild type, H168R has virtually no effect on the stability of R249S, but restores its binding affinity for the gadd45 promoter. NMR structural comparisons of R249S/H168R and R249S/T123A/H168R with wild type and R249S show that H168R reverts some of the structural changes induced by R249S. These results have implications for possible drug therapy to restore the function of tumorigenic mutants of p53: the function of mutants such as V143A and G245S is theoretically possible to restore by small molecules that simply bind to and hence stabilize the native structure, whereas R249S requires alteration of its mutant native structure.

Fig. 1. Urea-induced unfolding of human wild-type p53 and some mutant proteins. The experiments were performed at 10°C in 50 mM sodium phosphate and 5 mM DTT (pH 7.2). Protein concentration was 2 μM.

Fig. 2. Schematic model of the human p53 core domain tumour suppressor protein–DNA consensus sequence complex. The PDB (1tsr) co-ordinates were used to display only chain B of p53 (amino acids 94–312) in a complex with the DNA double helix (Cho et al., 1994). The mutated residues are shown in red and the zinc atom in blue. Figures 2 and 4 were made using the program MOLMOL (Koradi et al., 1996).

Fig. 3. Deviation of amide ¹H (A, C), ¹⁵N (B, D) chemical shifts from wild-type values. (A and B) Comparison of chemical shift changes between R249S and H168R/R249S. Residues that show recovery of wild-type chemical shifts upon the H168R mutation are indicated in (A) and (B) and colour coded red in Figure 4A. (C and D) Comparison of chemical shift changes between R249S/H168R and R249S/H168R/T123A. Residues with large chemical shift changes upon the third T123A mutation are indicated in (C) and (D) and colour coded red in Figure 4B. The chemical shifts for some residues were not determined (shown as gaps between data points) due to cross-peak overlapping or insufficient sensitivity in NMR experiments.

Fig. 4. Structural changes introduced by (A) the second suppressor mutation H168R and (B) the third mutation T123A. Chemical shifts of R249S, R249S/H168R and R249S/H168R/T123A were compared (see Figure 3).