Identification of protein-protein interaction networks by automated two-hybrid screening


As the human genome is unraveled, the focus of research is shifting to the functional analysis of gene products. In order to identify protein-protein interactions on a large scale we have developed an automated yeast two-hybrid system. We use interaction mating to generate large arrays of yeast clones containing protein-protein interactions. Furthermore, pipetting, picking and spotting robots are used for the parallel handling of large numbers of yeast clones.


Fig. 1: The automated yeast two-hybrid system

(A) Interaction mating in 384-well microtiter plates. (B) A protein-protein interaction network of Huntington's disease. (C) Pipetting robot for the pipetting of the baits to the preys and spotting robot for the gridding of yeast clones onto YPD-plates and onto high density membranes. (D) Growth of the plates and a ß-Galactosidase assay of high-density spotted nylon membranes.


Within the last two years we have applied the automated two-hybrid system for the identification of protein-protein interactions involved in HD. Using this method, a network of protein-protein interactions involved in Huntington's disease including 188 protein-protein interactions were found (Fig. 1, Fig. 2). Protein interactions were verified by in vitro binding experiments, co-immunoprecipitations and co-localization studies. The network permitted the functional annotation of 16 uncharacterized proteins and lead to the discovery of GIT1, a G protein-coupled receptor kinase interacting protein, crucial for Huntington's disease pathogenesis. Currently, the automated yeast two-hybrid system is used to identify the partner proteins of human disease proteins. We propose that the understanding of protein-protein networks of human diseases will help to identify novel targets for therapeutic intervention.



Fig. 2: View of a protein-interaction-network in Huntington's disease