Mapping protein‐protein interactions with combinatorial peptides

Brian K Kay

doi:10.1002/cfg.100

Abstract

Now that a number of genomes have been sequenced, attention has turned to understanding the complement of proteins encoded by the genome, which has been termed the ‘proteome’. One important aspect of analysis of the proteome is the identification of which proteins interact with each other; this information is invaluable in surmising the function of each protein and inferring what cellular pathway the protein may be a component of. While a variety of methodologies is widely used to map protein-protein interactions, such as affinity purification followed by mass spectrometry and yeast two-hybrid screening (both of which were covered at this meeting), an alternative approach is the use of phage-displayed combinatorial peptides. In phage-display, short oligonucleotides are inserted within a gene encoding a capsid (coat) protein of a bacteriophage, so that each viral particle displays a different peptide sequence. While it has been possible to clone and express short peptides attached to each of the five different capsid proteins of bacteriophage M13, the protein products of gene III and VIII are popular cloning sites for expression and display. Several excellent reviews of phage-display can be found elsewhere [6,22,5]. Over the past ten years, a considerable number of phage-displayed combinatorial peptide libraries have been generated [21]. With a library of recombinant phage particles in hand, it is possible to screen them by affinity selection with a variety of protein targets. Three rounds of selection (binding, elution, propagation) are typically sufficient to screen billions of different phage-displayed combinatorial peptides for those that bind to target proteins of interest. While there have been many recent publications of successful selection experiments, some of the more notable examples are the selection of peptide ligands which bind to receptors for erythropoietin [30], N-methyl D-aspartate (NMDA) [16], thrombopoietin [7], fibroblast growth factor [3], and estrogen [17]. When peptides are chemically synthesized corresponding to what is displayed by the selected phage, they generally bind with dissociation constant (Kd) values of 5 micromolar to 10 nanomolar to their cognate receptor, and typically have agonist or antagonist activities. Many different types of targets will yield phage after affinity selection, such as enzymes [12], growth factors [10,4], nucleic acids [2], and cells [20,13]. Comparative and Functional Genomics Comp Funct Genom 2001; 2: 304–306. DOI: 10.1002 / cfg.100

Highlights

That a number of genomes have been sequenced, attention has turned to understanding the complement of proteins encoded by the genome, which has been termed the ‘proteome’
While a variety of methodologies is widely used to map protein-protein interactions, such as affinity purification followed by mass spectrometry and yeast two-hybrid screening, an alternative approach is the use of phage-displayed combinatorial peptides
While there have been many recent publications of successful selection experiments, some of the more notable examples are the selection of peptide ligands which bind to receptors for erythropoietin [30], N-methyl D-aspartate (NMDA) [16], thrombopoietin [7], fibroblast growth factor [3], and estrogen [17]

Summary

Introduction

That a number of genomes have been sequenced, attention has turned to understanding the complement of proteins encoded by the genome, which has been termed the ‘proteome’. While a variety of methodologies is widely used to map protein-protein interactions, such as affinity purification followed by mass spectrometry and yeast two-hybrid screening (both of which were covered at this meeting), an alternative approach is the use of phage-displayed combinatorial peptides. With a library of recombinant phage particles in hand, it is possible to screen them by affinity selection with a variety of protein targets.

Results

Conclusion