Determining folding and binding properties of the C-terminal SH2 domain of SHP2

Abstract

SH2 domains are a class of protein-protein interaction modules with the function to recognize and bind sequences characterized by the presence of a phosphorylated tyrosine. SHP2 is a protein phosphatase involved in the Ras-ERK1/2 signaling pathway that possess two SH2 domains, namely, N-SH2 and C-SH2, that mediate the interaction of SHP2 with various partners and determine the regulation of its catalytic activity. One of the main interactors of the SH2 domains of SHP2 is Gab2, a scaffolding protein with critical role in determining cell differentiation. Despite their key biological role and the importance of a correct native fold to ensure it, the mechanism of binding of SH2 domains with their ligands and the determinants of their stability have been poorly characterized. In this article, we present a comprehensive kinetic study of the folding of the C-SH2 domain and the binding mechanism with a peptide mimicking a region of Gab2. Our data, obtained at different pH and ionic strength conditions and supported by site-directed mutagenesis, highlight the role of electrostatic interactions in the early events of recognition. Interestingly, our results suggest a key role of a highly conserved histidine residue among SH2 family in the interaction with negative charges carried by the phosphotyrosine of Gab2. Moreover, the analysis of the equilibrium and kinetic folding data of C-SH2 describes a complex mechanism implying a change in rate-limiting step at high denaturant concentrations. Our data are discussed under the light of previous works on N-SH2 domain of SHP2 and other SH2 domains.

Keywords: Chevron plot; Gab2; intermediate; kinetics; mutagenesis.

FIGURE 1

Panel (a): Equilibrium unfolding profiles of the C‐SH2 domain at different pH conditions. The dependence of the normalized fluorescence recorded at 330 nm as a function of the concentration of denaturant is reported. Panel (b): Urea‐induced denaturation performed at pH 8.0 and recorded at different wavelengths. In all cases data were globally fitted with a two state model equation, sharing the m _D–N value for all datasets (see details in the text)

FIGURE 2

Panel (a): Dependence of the logarithm of the observed rate constants for unfolding and refolding experiments as a function of [urea] at different pH conditions. Lines are the best fit to an equation describing a change in rate‐limiting step at high [urea] with the presence of a high‐energy intermediate along the reaction pathway. Panel (b): Analysis of observed initial (filled circles) and final (empty circles) fluorescence signal recorded in kinetic unfolding experiments at pH 5.5. The initial fluorescence, which resembles the fluorescence emission of the native state, show a linear dependence as function of [urea], indicating the absence of burst‐phase events

FIGURE 3

Panel (a): Pseudo‐first order kinetics of the binding reaction between dansyl‐Gab2 versus different concentrations of C‐SH2, at different pH conditions. As described in the text, points at 0 concentration of ligand were measured in separated displacement experiments. Panel (b): Pseudo‐first order kinetics of the binding reaction between dansyl‐Gab2 versus different concentrations of C‐SH2, at different ionic strength conditions (details of buffer used in Section 4 and in the legend). Panel (c): Dependence of logarithm of microscopic association (orange) and dissociation (green) rate constants as a function of pH. Lines are the best fit to Henderson–Hasselbalch equation. Panel (d): Dependence of logarithm of microscopic association (orange) and dissociation (green) rate constants and equilibrium dissociation rate constant (grey) as a function of ionic strength

FIGURE 4

Dependence of logarithm of microscopic association (orange) and dissociation (green) rate constants as a function of pH for H114A, H116A, H132A, H143A, and H196A variants. Kinetic parameters for H169A variant could not be measured (details in the text). Lines are the best fit to Henderson–Hasselbalch equation

FIGURE 5

Comparison of the three‐dimensional structure of the N‐SH2 domain (in blue) and C‐SH2 domain (in green) of SHP2. Reported structured are from PDB: 4qsy and PDB:4jeg, respectively. Structural alignment and corresponding image were performed produced with the UCSF Chimera software

FIGURE 6

Three‐dimensional structure of the C‐SH2 domain from PDB:4jeg with H114, H116, H132, H143, H169, and H196 in dark green color and ball‐and‐sticks format. In grey, it is reported a general ligand in order to highlight the position H169 residue in the binding pocket of the C‐SH2 domain