Prediction of protein-protein interaction sites using patch analysis¹

Abstract

A method for defining and analysing a series of residue patches on the surface of protein structures is used to predict the location of protein-protein interaction sites. Each residue patch is analysed for six parameters; solvation potential, residue interface propensity, hydrophobicity, planarity, protrusion and accessible surface area. The method involves the calculation of a relative combined score that gives the probability of a surface patch forming protein-protein interactions. Predictions are made for the known structures of protomers from 28 homo-dimers, large protomers from 11 hetero-complexes, small protomers from 14 hetero-complexes, and antigens from six antibody-antigen complexes. The predictions are successful for 66% (39/59) of the structures and the remainder can usually be rationalized in terms of additional interaction sites.

Introduction

The reliable prediction of protein-protein interaction sites is an important goal in the field of molecular recognition. It is of direct relevance to the design of drugs for blocking or modifying protein-protein interactions. Predictions can be divided into two main areas. The first is the docking of two proteins of known structure; a problem which has been addressed extensively using shape complementarity (e.g. Greer and Bush 1978, Wodak and Janin 1978, Kuntz et al 1982, Lee and Rose 1985, Connolly 1986, Jaing and Kim 1991, Helmer-Citterich and Tramontano 1994), chemical complementarity (e.g. Salemme 1976, Warwicker 1989) and combinations of the two (e.g. Walls and Sternberg 1992, Shoichet and Kuntz 1993, Vakser and Aflalo 1994). The second area of prediction, and the one addressed here, is the identification of putative interaction sites upon the surface of an isolated protein, known to be involved in protein-protein interactions, but where the structure of the partner or complex is not known.

It has been observed that protein-protein interaction sites in proteins have specific characteristics (e.g. Chothia and Janin 1975, Argos 1988, Janin et al 1988, Janin and Chothia 1990, Jones and Thornton 1995, Jones and Thornton 1996). In the accompanying paper (Jones & Thornton, 1997) we addressed the problem of comparing the observed interface with other similar sized patches on the protein surface using a series of parameters. It was concluded that it was possible to differentiate, to some degree, a protein interaction site from other similar patches on the surface of a protein. In the work presented here the use of multiple parameters for interface differentiation has been developed into a simple algorithm for the prediction of putative recognition sites for isolated proteins. Potentially this is a difficult problem, given that nothing is known about the partner protein. Therefore in this first attempt at prediction a relatively simple approach has been explored, to ascertain if prediction on this basis is possible. In this approach residue patches are defined on the surface of isolated proteins and analysed for a series of six parameters (solvation potential, residue interface propensity, hydrophobicity, planarity, protrusion and accessible surface area), with the aim of identifying those patches most likely to be involved in protein-protein interactions.