Intrinsic protein disorder and protein-protein interactions

Abstract

Intrinsically disordered proteins often bind to more than one partner. In this study, we focused on 11 sets of complexes in which the same disordered segment becomes bound to two or more distinct partners. For this collection of protein complexes, two or more partners of each disordered segment were selected to have less than 25% amino acid identity at structurally aligned positions. As it turned out that most of the examples so selected had similar 3D structure, the studied set was reduced to just these similar-fold cases. Based on the analyses of the interacting partners, the average sequence identity of the partners' binding regions showed substantially higher conservation as compared to the nonbinding regions: The residue identities, averaged over the 11 sets of partner proteins, were as follows: binding residues, 42 ± 6%; nonbinding residues 20 ± 3%; nonbinding buried residues 26 ± 5%; and nonbinding surface residues 16 ± 3%. The higher sequence identity of the binding residues compared to the other sets of residues provides evidence that these observed interactions are likely to be meaningful biological interactions, not artifacts. Since many of the features of the various interactions indicate that the disordered binding segments were likely to have been disordered before binding, these results also add further weight to the existence and function of intrinsically disordered regions inside cells.