Interaction of p53 with Mdm2 and azurin as studied by atomic force spectroscopy

Abstract

Azurin, a bacterial protein, can be internalized in cancer cells and induce apoptosis. Such anticancer effect is coupled to the formation of a complex with the tumour-suppressor p53. The mechanism by which azurin stabilizes p53 and the binding sites of their complex are still under investigation. It is also known that the predominant mechanism for p53 down-regulation implies its association to Mdm2, the main ubiquitin ligase affecting its stability. However, the p53/Mdm2 interaction, occurring at the level of both their N-terminal domains, has been characterized so far by experiments involving only partial domains of these proteins. The relevance of the p53/Mdm2 complex as a possible target of the anticancer therapies requires a deeper study of this complex as made up of the two entire proteins. Moreover, the apparent antagonist action of azurin against Mdm2, with respect of p53 regulation, might suggest the possibility that azurin binds p53 at the same site of Mdm2, preventing in such a way p53 and Mdm2 from association and thus p53 from degradation. By following the interaction of the two entire proteins by atomic force spectroscopy, we have assessed the formation of a specific complex between p53 and Mdm2. We found for it a binding strength and a dissociation rate constant typical of dynamical protein-protein interactions and we observed that azurin, even if capable to bind p53, does not compete with Mdm2 for the same binding site on p53. The formation of the p53/Mdm2/azurin ternary complex might suggest an alternative anti-cancer mechanism adopted by azurin.

Interaction of an anticancer peptide fragment of azurin with p53 and its isolated domains studied by atomic force spectroscopy

“…The relatively low unbinding frequency, detected even when specific interaction occurs, was thus to be ascribed to the random orientation resulting from the immobilization strategy. Interestingly, the unbinding frequency, observed to be consistent with values previously reported for other biological interactions, 27,39,40 was dependent on the loading rate, initially increasing, reaching a maximum, and decreasing as previously described. 27 As the unbinding force for each curve is the product of cantilever deflection (d, Figure 2) and its effective spring constant (k eff ), force histograms correspond to the different loading rates at which force spectroscopy measurements were conducted.…”

“…The cantilever nominal spring constant, k nom , was 0.5 N/m. The substrate was scratched to get a qualitative indication about the protein monolayer height as described by Funari et al 27 Force measurements were carried out in PBS buffer (50 mM K 3 PO 4 , 150 mM NaCl, pH 7.5) using force calibration mode AFM. The cantilevers used to perform force spectroscopy studies had a nominal spring constant, k nom , of 0.01 N/m.…”

“…A ligand-functionalized tip is approached to a surface covered by immobilized receptor (point 1); the cantilever begins to deflect upward due to the ligand-receptor intermolecular repulsive forces (point 2); the two partners can interact, and when the cantilever applies the maximum contact force upon the substrate, the approaching phase is stopped (point 3); the cantilever begins to retract, reaches the baseline deflection and begins to bend downward due to the attractive interaction force displayed by the ligand-receptor complex (point 4); when the force exerted by the cantilever overcomes the stability of the complex bonds, the cantilever jumps off, returning to its initial position (point 5). see Funari et al 27 for a detailed description.…”

“…The presence of a homogenous layer of p53 with discernible single proteins was observed. To verify the presence of a monolayer and define its height, a bare tip was thus used to scratch the substrate, working in contact mode as described in Funari et al 27 (results not shown). With a functionalized tip thousands of force curves were thus recorded on such a substrate with an intermediate degree of coverage (condition also suggested by Gilbert et al), 35 at several distinct points in which a corresponding percentage of specific events were observed.…”

“…This frequency could potentially be affected by the presence of unfavorable binding geometries and by steric hindrance. 16,27,38 However, studying the frequency of unbinding as a function of the surface coverage, no signif icant change was found. The relatively low unbinding frequency, detected even when specific interaction occurs, was thus to be ascribed to the random orientation resulting from the immobilization strategy.…”