Nat. Methods doi:10.1038/nmeth.2895

An estimated 20–30% of proteins in the human genome are membrane proteins, and many of them need to interact with other proteins to elicit their biological effect (or effects), for example, by initiating a signaling cascade in response to an extracellular stimulus. Identifying the proteins that are necessary and sufficient for the biological effect to manifest can be very challenging. Petschnigg et al. have modified the split-ubiquitin two-hybrid assay so that it can be used to detect interactions between a membrane protein and other proteins in mammalian cells. The authors showed that their method could be used to confirm that a GPCR–β-arrestin interaction only occurs in the presence of a small-molecule agonist and that a growth factor–adaptor protein interaction only takes place when the growth factor is phosphorylated at the proper position. They then determined that the assay could be scaled up; by screening 205 proteins that were computationally predicted to interact with EGFR, they identified 124 putative EGFR binders, many of which were previously known to bind EGFR or subsequently shown, in follow-up experiments, to interact with this growth factor. The authors believe that their method could be used to help unravel large-scale protein-protein interaction networks and facilitate the development of drugs for diseases that are caused by the mutation or dysregulation of membrane proteins.