Biological systems function through interactions of proteins with other macromolecules (proteins, DNA, RNA, complex carbohydrates and lipids), and with small molecules, including metabolites and secondary compounds. Membranes provide a surface for perceiving and transducing signals from adjacent cells and external conditions such as availability of nutrients, or toxins and adverse conditions for growth and development. In addition, membranes control cellular and subcellular entry and exit of molecules. To understand the regulation of cell-environment, cell-cell, and intracellular compartment to cytosol interactions better, we plan to determine the interactions of 8,400 proteins from the reference plant Arabidopsis. This set comprises most integral membrane proteins (~4,300) and a large number (>1,000) of proteins predicted to be involved in signaling or protein modification.

Since membrane proteins are inherently difficult to work with, a special yeast two-hybrid system, the split-ubiquitin system, was developed for membrane protein interactions. In this project, a systematic analysis of the binary interactions of these >6000 proteins will be performed using the split ubiquitin system, culminating in tests of more than 25 million interactions (assuming an ~80% success rate in cloning and yeast matings). A subset of interactions will be verified using the split GFP system.

The data will be subjected to bioinformatic and graph-theoretical meta-analysis to identify the main signal transduction pathways and protein complexes (signalosomes) in the interaction network. Key functional signal mediators (hubs) as well as abundant regulatory patterns will be identified. T-DNA insertional knockouts of genes encoding 80 proteins comprising critical nodes for maintaining the structure and/or function of the predicted pathways and complexes will be characterized for macroscopic phenotypes, and a smaller set of 20, predicted to be expressed in guard cells or the root epidermis from previous microarray analyses, will receive detailed characterization regarding their functions in those cell types, reflecting the specific expertise areas of the PIs.

Data will be made public through our project website, TAIR, and interactions data repositories such as IntAct, DIP and bioGRID. The project will provide an additional source of full length ORFs and will contribute significantly to the assignment of functions to proteins; for example 37% of the membrane proteins in our dataset are currently annotated as unknowns in one or both of the categories of 'function' and 'process'. Moreover, since these data will provide the first complete set of membrane protein interactions among each other and the signaling proteome in any organism, they will provide a resource for generation of novel hypotheses, pursuit of biotechnological applications, and development of comparative genomics studies.

The minority-serving schools San JoseU State University, City College of San Francisco, San Francisco State University, historically black universities, and Swarthmore College will be targeted as sources of undergraduates who will select and assess specific signaling modules within our interactome using wet bench, bioinformatic and theoretical tools. High school students from the Preuss School in San Diego and the Bald Eagle Area High School near Penn State, schools with a large proportion of economically disadvantaged students, will be selected to participate in identification and phenotypic analysis of T-DNA insertional mutants. It is hoped that genuine participation in a research project will encourage these students to consider careers in science. In addition, the project will train participating graduate students in interdisciplinary research, specifically the concepts and tools of systems biology.