A microarray-based approach to evaluate the functional significance of protein-binding motifs.

Intracellular proteins comprise numerous peptide motifs that interact with protein-binding domains. However, using sequence information alone, the identification of functionally relevant interaction motifs remains a challenge. Here, we present a microarray-based approach for the evaluation of peptides as protein-binding motifs. To this end, peptides corresponding to protein interaction motifs were spotted as a microarray. First, peptides were titrated with a pan-specific binder and the apparent K(d) value of this binder for each peptide was determined. For phosphotyrosine-containing peptides, an anti-phosphotyrosine antibody was employed. Then, in the presence of the pan-specific binder, arrays were competitively titrated with cell lysate and competition constants were determined. Using the Cheng-Prusoff equation, binding constants for the pan-specific binder and inhibition constants for the lysates were converted into affinity constants for the lysate. We experimentally validate this method using a phosphotyrosine-binding SH2 domain as a further reference. Furthermore, strong binders correlated with binding motifs engaging in numerous interactions as predicted by Scansite. This method provides a highly parallel and robust approach to identify peptides corresponding to interaction motifs with strong binding capacity for proteins in the cell lysate.

A peptide-functionalized polymer as a minimal scaffold protein to enhance cluster formation in early T cell signal transduction.

In cellular signal transduction, scaffold proteins provide binding sites to organize signaling proteins into supramolecular complexes and act as nodes in the signaling network. Furthermore, multivalent interactions between the scaffold and other signaling proteins contribute to the formation of protein microclusters. Such microclusters are prominent in early T cell signaling. Here, we explored the minimal structural requirement for a scaffold protein by coupling multiple copies of a proline-rich peptide corresponding to an interaction motif for the SH3 domain of the adaptor protein GADS to an N-(2-hydroxypropyl)methacrylamide polymer backbone. When added to GADS-containing cell lysates, these scaffolds (but not individual peptides) promoted the binding of GADS to peptide microarrays. This can be explained by the cross-linking of GADS into larger complexes. Furthermore, following import into Jurkat T cell leukemia cells, this synthetic scaffold enhanced the formation of microclusters of signaling proteins.

Multivalent presentation of the cell-penetrating peptide nona-arginine on a linear scaffold strongly increases its membrane-perturbing capacity.

Arginine-rich cell-penetrating peptides (CPP) are widely employed as delivery vehicles for a large variety of macromolecular cargos. As a mechanism-of-action for induction of uptake cross-linking of heparan sulfates and interaction with lipid head groups have been proposed. Here, we employed a multivalent display of the CPP nona-arginine (R9) on a linear dextran scaffold to assess the impact of heparan sulfate and lipid interactions on uptake and membrane perturbation. Increased avidity through multivalency should potentiate molecular phenomena that may only play a minor role if only individual peptides are used. To this point, the impact of multivalency has only been explored for dendrimers, CPP-decorated proteins and nanoparticles. We reasoned that multivalency on a linear scaffold would more faithfully mimic the arrangement of peptides at the membrane at high local peptide concentrations. On average, five R9 were coupled to a linear dextran backbone. The conjugate displayed a direct cytoplasmic uptake similar to free R9 at concentrations higher than 10µM. However, this uptake was accompanied by an increased membrane disturbance and cellular toxicity that was independent of the presence of heparan sulfates. In contrast, for erythrocytes, the multivalent conjugate induced aggregation, however, showed only limited membrane perturbation. Overall, the results demonstrate that multivalency of R9 on a linear scaffold strongly increases the capacity to interact with the plasma membrane. However, the induction of membrane perturbation is a function of the cellular response to peptide binding.

Peptide microarrays to probe for competition for binding sites in a protein interaction network.

Cellular protein interaction networks are a result of the binding preferences of a particular protein and the entirety of interactors that mutually compete for binding sites. Therefore, the reconstruction of interaction networks by the accumulation of interaction networks for individual proteins will greatly overestimate connectivity within the network. Here, we addressed the impact of intracellular complexity on signalling networks using microarrays that carried a collection of peptides binding to the GRB2 SH2 and SH3 domains. Binding patterns and affinities for the recombinant adaptor protein GRB2 were compared with the ones for the protein in cell lysates. Peptide microarrays were titrated with the histidine-tagged recombinant protein, cell lysates or mixtures of both. Indeed, for recombinant GRB2, binding was detected for more peptides than for GRB2 in cell lysates. Moreover, binding was also observed for poor binders. It was impossible to define affinity thresholds for the binding of the recombinant protein to enable a discrimination of physiologically relevant interactions. Titrations of recombinant protein with lysate confirmed competition as the basis for fewer interactions. Importantly, the methods presented here enable the description of physiologically relevant binding patterns for proteins of interest and the identification of those peptide motifs, which are most strongly affected by competition. BIOLOGICAL SIGNIFICANCE: The biological significance of protein-protein interactions can only be addressed in a physiologically meaningful way in the presence of the endogenous proteome which may contain proteins that compete for binding sites. Using peptide microarrays, we here demonstrate for the adaptor protein GRB2 that this competition strongly reduces the number of interactions with other signalling proteins.

Analyzing the homeostasis of signaling proteins by a combination of Western blot and fluorescence correlation spectroscopy.

The determination of intracellular protein concentrations is a prerequisite for understanding protein interaction networks in systems biology. Today, protein quantification is based either on mass spectrometry, which requires large cell numbers and sophisticated measurement protocols, or on quantitative Western blotting, which requires the expression and purification of a recombinant protein as a reference. Here, we present a method that uses a transiently expressed fluorescent fusion protein of the protein-of-interest as an easily accessible reference in small volumes of crude cell lysates. The concentration of the fusion protein is determined by fluorescence correlation spectroscopy, and this concentration is used to calibrate the intensity of bands on a Western blot. We applied this method to address cellular protein homeostasis by determining the concentrations of the plasma membrane-located transmembrane scaffolding protein LAT and soluble signaling proteins in naïve T cells and transformed T-cell lymphoma (Jurkat) cells (with the latter having nine times the volume of the former). Strikingly, the protein numbers of soluble proteins scaled with the cell volume, whereas that of the transmembrane protein LAT scaled with the membrane surface. This leads to significantly different stoichiometries of signaling proteins in transformed and naïve cells in concentration ranges that may translate directly into differences in complex formation.