Über die Autorin bzw. den Autor

ERNEST GIRALT, Department of Organic Chemistry, University of Barcelona and Institute for Research in Biomedicine, Barcelona, Spain

MARK PECZUH, Department of Chemistry, University of Connecticut, USA

XAVIER SALVATELLA, ICREA and Institute for Research in Biomedicine, Barcelona, Spain

Von der hinteren Coverseite

Protein Surface Recognition: Approaches for Drug Discovery

A new perspective on the design of molecular therapeutics is emerging. This new strategy emphasizes the rational complementation of functionality along extended patches of a protein surface with the aim of inhibiting protein-protein interactions (PPIs). The successful development of compounds able to inhibit these interactions offers a unique chance to selectively intervene in a large number of key cellular processes related to human disease.

Protein Surface Recognition: Approaches for Drug Discovery presents a detailed treatment of this strategy. Starting with a survey of PPIs that are key players in human disease and biology and the potential for therapeutics derived from this new perspective, the book then examines the fundamental physical issues that surround protein-protein interactions that must be considered when designing ligands for protein surfaces. Examples of protein surface-small molecule interactions, including treatments of protein-natural product interactions, protein-interface peptides, and rational approaches to protein surface recognition are given. Finally, the book surveys techniques that will be integral to the discovery of new small molecule protein surface binders, from high throughput synthesis and screening techniques to in silico and in vitro methods for the discovery of novel protein ligands, and ends with two extended case studies - inhibitors of the MDM2-p53 PPI, and the discovery of potent LFA-1 antagonists.

Protein Surface Recognition: Approaches for Drug Discovery provides an intellectual ‘tool-kit' for investigators in medicinal and bioorganic chemistry looking to exploit this emerging paradigm in drug discovery.

Auszug. © Genehmigter Nachdruck. Alle Rechte vorbehalten.

Protein Surface Recognition

John Wiley & Sons

Chapter One

C. W. Bertoncini, A. Higueruelo and X. Salvatella

1.1 Introduction

The regulation of protein–protein interactions (PPIs) is fundamental for cellular function because PPIs are involved in virtually all biological processes. A complete and detailed description of the interaction map for proteins, known as interactome, is therefore one of the most important challenges in molecular biology, one that will provide great opportunities for therapeutic intervention in the complex diseases that challenge the biomedical community and the pharmaceutical industry. In this chapter we provide an overview of the different techniques that are currently available for the discovery and structural and thermodynamic analysis of PPIs as well as a survey of the general structural and dynamical properties of proteins and protein complexes that affect drug design. Rather than a comprehensive survey of the technical literature on methods to screen and characterize PPIs we present here a general discussion of these tools and refer the reader to the reviews and examples of application that we cite to identify the primary literature.

1.2 Techniques to Identify Protein-Protein Interactions

Many methods have been developed for the isolation and characterization of protein complexes, both in vitro and in vivo. Among them five methodologies are particularly suitable for high-throughput, and account for the majority of proteome-wide studies.

1.2.1 The Yeast Two Hybrid Assay (Y2H)

This system exploits the formation of a stable complex between interacting proteins to bring together two modules of a cis-acting transcriptional promoter, stimulating the expression of a reporter gene. It requires the construction of two hybrid genes, one encoding the DNA-binding domain (BD) of the transcription factor fused to a target protein (the bait) and a second encoding its transcription-activation domain (AD) fused to a different protein (the prey). If the prey and bait proteins interact through a PPI the two modules of the transcription factor (BD and AD) are brought together to reconstitute the transcription activity. Provided that the interaction between the prey and bait proteins is sufficiently strong, the now functional transcription factor will bind to the promoter sequence in the proximity of the reporter gene, via its DNA binding domain (BD), and recruit the transcriptional machinery, via its transcription-activation domain (AD, Figure 1.1A). The most commonly employed DNA-binding domains are derived from the yeast Gal4 and LexA transcription factors, while activating domains come also from Gal4 or from the viral activator VP16. Expression of the reporter gene gives the yeast a unique characteristic which allows identification of a successful PPI interaction between the bait and prey proteins. Reporter genes commonly employed are lacZ, that codifies for the enzyme β-galactosidase, that metabolizes X-gal (5-bromo-4-chloro-3-indolyl-β-D-galactoside) to give a distinctive blue color, or auxotrophic genes such as HIS3, LEU2 or URA3, which confer positive colonies the ability to growin media lacking specific nutrients.

One key advantage of the Y2H assay is its in vivo nature, that allows the investigation of PPIs under physiological conditions. Additional advantages of this method are its high sensitivity—it can detect very weak and therefore transient interactions, with Kd as low as 10^-7 M—its scalibility and its easy automation. The Y2H assay can also be used in a quantitative fashion to determine the strength of the interaction between the bait and prey proteins by monitoring the amount of reporter protein produced, for example by measuring, when using the lacZ reporter gene, the β-galactosidase activity. The main disadvantage of the Y2H approach to the identification of PPIs is the number of control experiments that it requires, that are mainly aimed at determining whether the bait and prey proteins have affinity for DNA and are indeed capable of self-activating the transcription of the reporter. An additonal concern when using this approach, one that is directly linked to the its in vivo nature, is the possibility that a third protein mediates, in the assay, the interaction between the pray and bait proteins; it is therefore important to validate all PPIs derived from this assay by other methods, including those discussed in this chapter. Other limitations of the Y2H assay are due to its use of yeast, as some post-translational modifications are different in yeast to those in other eukaryots, and to the localization of of interactions in the nucleus, where some target PPIs may experience an incorrect cellular environment. Membrane proteins are obviously not suitable for this assay, but interactions between the cytoplasmic domains of extracellular receptors can be screened, using this approach, to study signal transduction pathways. PPIs identified from genomic-scale Y2H analysis are expected to have a success rate of 50%, and bioinformatic analysis tools to refine the results with co-expression and co-localization analysis can very significantly increase the accuracy of the results.

1.2.2 Phage Display

This method was one of the earliest tools developed for screening PPIs, before the recent spread of mass spectrometry-assisted protein identification. Phages are bacteria-specific viruses which carry the viral DNA enclosed in an envelope of viral proteins. Phage particles are therefore unique in that they contain both DNA and protein copies of a given gene in a single entity. This singularity provided molecular biologist with a unique tool to isolate simultaneously both the protein displayed in the exterior of a phage particle and its DNA sequence.

The phage display technique involves the construction of a DNA library where the sequences that code for the proteins to be screened are fused to the sequence of a bacteriophage coat protein (P8 or P3) in a plasmid containing the rest of the components of the phage genome (6.5 Kbp for the filamentous bacteriophage M13). Upon infecting an E. coli host the phages display the chimeric proteins in their outer surface and bear inside the DNA sequences that correspond to such proteins. Phages are produced in E.coli individually, and the particles are then assayed for binding to the immobilized target protein in an ELISA (Enzyme-Linked Immunsorbent Assay) fashion. In order to reduce background bound phages are collected and re-amplified in E. coli and, after two to three rounds of binding, the DNA of phages strongly interacting with the target is isolated and sequenced, leading to the identification of the proteins interacting with the target protein. A concise recollection of several random peptide and genomic libraries constructed in different phage vectors, as well as different kind of proteins successfully displayed infilamentous phages was published in 1997 by Smith and Petrenko. Despite being slightly outdated, this survey highlights the range of proteins that withstand phage display that includes enzymes, hormones, receptors, cytokines and DNA binding proteins and account for more than 50 publications.

The range of target molecules that have been subjected to phage display-based screening is very wide, and is by no means restricted to polypeptides. Smith and Petrenko also collected published data on the range of target proteins that have been screened, that includes antibodies, Calmodulin, the tumor suppressor P53, Hsc70, integrins and hormones. Three applications are worth mentioning in the context of PPIs:

(i) the identification of epitopes to monoclonal antibodies, by constructing naive phage peptide libraries of 10 to 40 amino acids;

(ii) the analysis of interfaces at the residue level by alanine scanning mutagenesis;

(iii) the high-throughput determination of surfaces and free energies of binding];

(iv) the identification and construction of new scaffolds for PPIs, like single domain β-sandwich proteins (FN3 and V_HH), ankyrin repeats, WW or SH3 domains, and four helix boundels.

Phage display is not, however, without disadvantages. An important one is, for certain applications, the need for the constructed library to be a representative sample of the whole genome, that may be challenging for some laboratories; the libraries can, however, now be obtained commercially, and once produced, can be replicated by passage through a bacterial host. An additional major limitation arises when the displayed proteins fold rapidly because the chimeric fusion protein needs to be remain unfolded for efficient secretion to the bacterial periplasm, prior to display; this has however been recently overcome by the use of alternative translocation pathways and by signal recognition particles. A third major potential problem concerns the immobilization of the target protein, which may hinder the interaction surface: GST or His tags are therefore desirable to aid in immobilization.

1.2.3 Protein Microarrays

A recently proposed method to analyse PPIs on a genomic scale uses functional protein microarrays, where thousands of recombinantly expressed and purified proteins are individually spotted on a surface, by chemical derivatization, to constitute the panel of proteins to be screened (Figure 1.1B). A single fluorescently labelled protein (or ligand), is then put in contact with the array in buffered aqueous solution, and subsequently washed with incrementing stringency. Following this the microarray slide is read by a scanner with laser excitation and fluorescence detection capabilities to identify fluorescent spots indicative of the occurrence of a PPI between the labelled ligand and a protein of the micro array. An important advantage of this method is that variable solution conditions can be easily assayed; this makes it possible, for example, to characterize binding at different concentrations in a high-through put fashion to report on the thermodynamic stability of the PPIs detected. This approach has been successfully employed to identify and characterize proteins interacting with the Erb recept or family, where affinities using the protein microarray where comparable to those determined by surface plasmon resonance.

The main advantage of this technique lies on its ability to screen thousand of interactions simultaneously on a single chip. However care must be taken when interpreting some interactions, in particular low affinity ones, as the chemical derivatization process may affect the properties of immobilized proteins. In addition, checks for correct expression and adequate immobilization have to be carried out; for this purpose and it is common for proteins to carry an extra peptide tag which allows identification in western blots and in the microarray slide.

1.2.4 Affinity-based Methods

A number of methods have been developed to specifically isolate protein complexes formed in vivo and further analyse them by mass spectrometry. The main idea is to fuse the protein of interest (bait) to a peptide tag which confers affinity to a ligand immobilized on a solid support. Proteins that establish a PPI with the tagged protein can in this way be co-isolated upon incubation with the ligand matrix, and complexes can then be eluted by incubation with free ligand. Modern MS methodologies are key in this approach as they are used to analyse the bound proteins.

The general procedure involves the construction of the gene for the chimera that fuses the coding sequence of the bait to the desired tag. The plasmid is then transfected into a eukaryotic cellular host, where it is expressed, producing large amounts of the protein. Cells are then lysed, and the lysates are subjected to affinity chromatography, where protein complexes involving the tagged bait are specifically isolated. Proteins composing the complexes are resolved by polyacrilamide gel electrophoresis (SDS-PAGE), bands are excised and then subjected to tryptic digestion to produce peptides suitable to MS analysis. Such standard methodologies include the use of Matrix Assisted Laser Desorption Ionization (MALDI) MS, or liquid chromatography coupled to Electro Spray Ionization (ESI) MS.

This is a simple methodology that is recommended for most laboratories, as it is relatively inexpensive, requires neither complex equipment nor commercial services and can be carried out with the help of commercial kits. Affinity-based methods normally identify high affinity interactions i.e. with slow kinetics of dissociation, and one of their great advantages is that they allow the isolation of multiprotein complexes, that is not possible when using the Y2H or phage display assays or protein microarrays. It is however important to acknowledge that the use of a peptide tag can promote or impair certain PPIs, affect the normal localization of the bait protein as well as impair the isolation of the protein–protein complexes if the tag becomes buried as they form. All these problems are, however, easily overcome by the use of a second unrelated tag in further similar experiments aimed at confirming the PPI. Depending on the nature of the tag, it is useful to classify affinity methods in three groups:

1.2.4.1 Single Tag Affinity Purification

This method involves the use of a unique peptide motif at the N- or C-terminus of the bait protein to detect protein–protein interactions that occurr in vivo by co-sedimentation of the interacting partners. One of the most extended methodologies involves the use of the gluthatione-S-transferase (GST)protein as a fusion of one of the assayed proteins. The GST tag confers the bait protein high affinity to gluthatione, which is immobilized on agarose beads to pull down interacting proteins from cellular extracts. The disadvantages of this technique lie in the considerable size of GST (27 KDa) that can perturb the structure of the fused protein, and in the co-isolation of proteins interacting with GST itself rather than with the bait. Similar approaches employ a poly-Histidine tagged protein with high affinity to metal-chelated beads; this tag only slightly perturbs the structure of proteins, but usually results in the isolation of His-rich proteins that are false positives. Other motifs widely used as tags include maltose binding protein (MBP), immunoglobulin binding domains (protein A or G), and the Strep-tag, which is based on the high affinity biotin/ streptavidin interaction.

1.2.4.2 Tandem Affinity Purification (TAP)

TAP is a modified version of the single tag affinity method, and involves two different peptide motifs in tandem, separated by a protease cleavage site (Figure 1.2B). The improvement in TAP in respect to the single tag method lies on the usage of two affinity purification steps which reduces the presence of spurious interacting proteins that can lead to false positives. The initial combination of tags featured a tandem of protein A and a Calmodulin binding peptide (CaMBP), separated by a Tobacco Etch Virus (TEV) protease cleavage site. Protein complexes involving the tagged-bait protein are first isolated with immunoglobulin-agarose beads that have high affinity for protein A. Digestion with TEV protease releases the complex and exposes the Calmodulin Binding Peptide (CaMBP). Incubation with CaM-coated beads followed by and elution with EGTA or free CaMBP allows isolation of the purified complex. A new generation of tags involves high efficiency cloning vectors, the use of inducible promoters of expression, tetracysteine motifs suitable for in cell fluorescence imaging, and streptavidin tags. Proteome wide scale studies in yeast by the TAP method have recently identified more than 500 protein complexes of physiological relevance, demonstrating the high-throughput capabilities of the technique.

1.2.4.3 Co-immunoprecipitation (Co-IP)

A protein complex stabilized by PPIs present in a cellular or tissue homogenate can be isolated by means of an appropriate antigen-antibody pair followed by affinity chromatography with protein A or G-coated beads. Antibodies suitable for Co-IP studies can be raised against the protein of interest, or against a small peptide tag fused to the protein of interest; the second option is preferable since it ensures the absence of cross reactions and allows the use of already characterized commercial antibodies. Commonly employed tags for Co-IP studies are HA ([MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]), c-Myc ([MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]), FLAG ([MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]), all of which have plasmids and antibodies commercially available.

(Continues...)

Excerpted from Protein Surface Recognitionby Ernest Giralt Mark Peczuh Xavier Salvatella Copyright © 2011 by John Wiley & Sons, Ltd. Excerpted by permission of John Wiley & Sons. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.