Interactive proteomics addresses the physical associations among proteins and establishes global, disease-, and pathway-specific protein interaction networks. The inherent chemical and structural diversity of proteins, their different expression levels, and their distinct subcellular localizations pose unique challenges for the exploration of these networks, necessitating the use of a variety of innovative and ingenious approaches. Consequently, recent years have seen exciting developments in protein interaction mapping and the establishment of very large interaction networks, especially in model organisms. In the near future, attention will shift to the establishment of interaction networks in humans and their application in drug discovery and understanding of diseases. In this review, we present an impressive toolbox of different technologies that we expect to be crucial for interactive proteomics in the coming years.

Introduction

Proteins are the basic building components of the cell that carry out diverse functions such as catalyzing chemical reactions as enzymes, binding other molecules as receptors, and forming structural components. Physical associations between different proteins, so-called protein-protein interactions (PPIs), are an essential aspect in all biological pathways and signaling mechanisms, and are strongly predictive of functional relationships. Mapping of PPI networks is particularly useful for the functional annotation of the large number of newly identified open reading frames (ORFs) that were identified in the recent sequencing of the human genome and the genomes of various model organisms. The exploration of PPIs on a large-scale or global level is referred to as interactive proteomics, cell-map, or interaction proteomics (1,2). Analysis of protein assemblies can also be considered as one part in the wider context of structural proteomics, which is the systematic analysis of protein structure and folding (3,4). Establishing networks of PPIs will eventually lead to a better understanding of cellular pathways and functional associations and allow a more holistic view of the cell, as opposed to the classical one-protein-at-a-time approach.

Indeed, the recent exploration of entire interactomes from different organisms was a consequence of the development of technologies that allowed the assessment of PPIs on a high-throughput (HTP) level. In this review, we begin with the current state of interactive proteomics research, describing established mature technologies like yeast two-hybrid (Y2H) and affinity purification-mass spectrometry (AP-MS), which have been applied for genome-wide or pathway-specific screens in different model organisms and in human cells. We then present emerging technologies, including in vivo methodologies (Y2H derivatives, protein fragment complementation in mammalian cells, detection of PPIs by fluorescence and bioluminescence) and biochemical approaches (MS-based methods, large-scale analysis of tagged arrays). We discuss these methodologies in the context of future directions in interactive proteomics, considering the extension of large-scale approaches into mammalian organisms, and the application of interactive proteomics for drug discovery and research on human diseases. Unfortunately, space restrictions leave us unable to discuss technologies that are currently used to explore PPIs in vitro. This includes the protein-chip technology that has emerged as a powerful tool for the parallel analysis of protein interactions and biochemical activities (5,6). Likewise, we refer the reader elsewhere for excellent reviews on protein engineering and phage display (7,8,9). We will also not discuss data interpretation and the analysis of network topologies that are addressed by bioinformatics (10,11).

Genetic Methods for the Analysis of Protein Interactions

Yeast Two-hybrid Assay

In the time since Song and Fields devised their original interaction trap assay in 1989 (12), the Y2H system has become the most widely used method to assess both individual PPIs and entire interactomes. Y2H is very cost effective, convenient to use, and easily adaptable for HTP screening procedures. Exploiting the modularity of eukaryotic transcription factors, a known protein (bait) protein is fused to the DNA-binding domain (DBD), and an interacting protein (prey) is fused to the activation domain (AD) of a transcriptional activator ((Figure 1)A). A physical interaction between bait and prey leads to the activation of a reporter gene construct that serves as a selection marker. Using the yeast transcription machinery as a vehicle, physical interactions between proteins from any other organism can be easily monitored. However, a major drawback in Y2H is false-positive interactions that occur upon self-activation of reporter genes by individual bait and prey proteins. Furthermore, since interactions between all bait and prey proteins are confined in the nucleus of a lower eukaryote, PPI analysis by Y2H suffers from a lack of contextual specificity. Consequently, a variety of methods have been developed to improve the Y2H system and adapt it for different purposes. For more reading on Y2H, we refer to the article of Ratushny and Golemis in this issue of BioTechniques (p. 655) and other reviews (13,14,15).

Figure 1.

Several years ago, two different laboratories generated a comprehensive overview of the yeast Saccharomyces cerevisiae proteome by Y2H, using matrixes with all 6000 yeast ORFs cloned into bait and prey vectors (16,17). Since then, Y2H screens have been done for the metazoan model organisms Drosophila melanogaster and Caenorhabditis elegans (18,19) and, more recently, for Homo sapiens (20,21). Because of the sheer size and complexity of the mammalian proteomes, human Y2H screens are more practical on a subgenome scale and have been centered upon specific signaling pathways or disease factors (22,23,24). In one of these studies, Y2H screens were performed with bait proteins that are involved in several known inherited ataxias (23). Inherited ataxias are neurodegenerative disorders with common clinical and pathological features. The molecular basis of these diseases, however, is not well defined. Interestingly, Y2H screens revealed that many ataxia-associated proteins share interacting partners, suggesting that these proteins are functionally connected and act in a concerted manner. This work provides an excellent understanding of how interactome studies using Y2H can help us understand common pathogenic mechanisms and identify candidate genes for complex human diseases.