Why Map interactomes?
Protein-protein interactions are essential to the functioning of cells. As such, protein-protein interaction maps (interactome maps) are indispensable for systems biology and focused biological studies, because:
How to map interactomes?
Three fundamentally disparate strategies have been used to map interactome networks (Figure): i) curation of protein interaction data in the scientific literature; ii) computational predictions of protein interactions based on available orthogonal information such as sequence similarity or co-presence of genes in sequenced genomes; and iii) systematic, unbiased high-throughput experimental mapping strategies applied at proteome-scale. These complementary approaches differ in the value of the information they contain. Literature curated interactome maps comb through published experimentally derived information, but are limited by the inherently variable quality of curation (Cusick et al, Nat Methods 2009). “Celebrity proteins” are studied more frequently than others, resulting in a bias towards higher connectivity compared to other proteins (Yu et al, Science 2008). Computational predictions can be applied at genome scale for moderate cost, but are limited by current understanding of biological systems and even then should be tested experimentally before firm conclusions can be drawn (Carvunis et al, Handbook Syst Biol 2012). High-throughput experimental approaches generate unbiased, systematic and well-controlled data. They either test all binary combinations of possible protein pairs to determine which ones interact directly (Braun P, Proteomics 2012), or identify protein membership of protein complexes isolated from cells, that is, indirect interactions (Gingras & Raught, FEBS Lett 2012). Mapping of the binary interactome is carried out at CCSB primarily by updated variants of yeast two-hybrid methodologies (Chien et al, PNAS 1991).
CCSB has released high-confidence binary protein interaction networks for several organisms (Figure below), based on robust high-throughput yeast two-hybrid (Y2H) technology (Dreze et al, Methods Enzymol 2010). We have reported first draft binary interactome maps for H. sapiens (Rual et al, Nature 2005) and C. elegans (Walhout et al, Science 2000; Reboul et al, Nat Genet 2003; Li et al, Science 2004) and A. thaliana (Arabidopsis Interactome Mapping Consortium, Science 2011), as well as second generation interactome maps for S. cerevisiae (Yu et al, Science 2008) and C. elegans (Simonis et al, Nat Methods 2009) and H. sapiens (Rolland et al, Cell 2014).
Interactome mapping requires constant betterment of technologies and strategies to increase throughput, to improve quality, and to decrease cost. Our proven empirical framework for binary interactome mapping (Figure below) can provide high-confidence, validated interaction data for any binary interactome mapping project (Venkatesan et al, Nat Methods 2009; Braun et al, Nat Methods 2009; Cusick et al, Nat Methods 2009). Greater throughput comes through application of next generation sequencing technologies to generate interactome datasets (Yu et al, Nat Methods 2011).
A second-generation binary interactome map for H. sapiens has been published (Rolland et al, Cell 2014), with the data available in the CCSB Human Interactome Database. Current human interactome mapping efforts take advantage of a greatly expanded human ORF collection and more robust pipeline strategy. Such efforts are supported by NIH grant U01-HG001715 “Mapping the first half of the reference human binary protein interactome”.
A next generation interactome map for yeast S. cervisiae is underway, exploring a diverse array of Y2H assays (Chen et al, Nat Methods 2010), and testing an expanded yeast ORFeome incorporating many proto-gene predictions (Carvunis et al, Nature 2012). Yeast interactome mapping efforts are supported by NIH grant R01-HG006061 ”An S. cerevisiae high-coverage high-quality protein-protein binary interactome map”.
A next generation interactome map for the fly D. melanogaster is beginning, exploiting the availability of fly ORFeome resources at the Berkeley Drosophila Genome Project. Fly interactome mapping efforts are supported by NIH grant R01-HG007118 "Large-scale high-confidence binary protein interaction network for Drosophila", in collaboration with the Perrimon lab at Harvard Medical School and the Celniker lab at Lawrence Berkeley National Laboratory.