els a0022595.tex V2 - 03/23/2017 6:49 P.M. Page 1 Version 2 a0022595 Protein–Protein Interactions: The Structural Foundation of Life Complexity Luisa Di Paola, Unit of Chemical-physics Fundamentals in Chemical Engineering, Department of Engineering, Università Campus Bio-Medico di Roma, Rome, Italy Alessandro Giuliani, Environment and Health Department, Istituto Superiore di Sanità, Roma, Italy Based in part on the previous version of this eLS article ‘Protein–Protein Interactions’ (2001) by Jeremy H Lakey and Isa Gokce. AU:1 Introductory article Article Contents • Setting the Frame • Methods to Investigate Protein–Protein Interactions • Perspective: The Network Approach to PPI • Applications Online posting date: 15 th May 2017 The amazing variety of protein functions are often covered by protein complexes, so understanding protein–protein interactions means coming deeply into the functional role of proteins in life. In the last years, the investigation of protein–protein interactions has become central in many fields, spanning from molecular biology to pharmacology. In this article, we present a state of the art of methods for such investigation, along with perspectives of applications. We stressed the multiscale nature of approaches, longing from genome-wide analysis to the detailed study of protein–protein interface on single residues. The most innovative approaches, based on com- plex network theory, shed a very bright light on future trends for protein–protein applications on drug design and on molecular therapy for diseases where protein association plays a pivotal role (mis- folding). Setting the Frame From introductory physical-chemistry handbooks, we know that AU:2 a triatomic ordered collision is extremely rare (nearly impossible) in diffusive regime. Shifting to biochemistry, we learn about ordered sequences of chemical reactions involving dozens of ordered collisions. These long pathways are located at the very heart of life involving energy production (e.g. Krebs cycle), eLS subject area: Biochemistry How to cite: Paola, Luisa Di and Giuliani, Alessandro (May 2017) Protein–Protein Interactions: The Structural Foundation of Life Complexity. In: eLS. John Wiley & Sons, Ltd: Chichester. DOI: 10.1002/9780470015902.a0001346.pub2 complex organic syntheses (e.g. lipid metabolism) and signalling (e.g. apoptosis). Reconciliation between the physical-chemistry laws and the striking order of cell metabolism is only possible taking into account protein–protein interactions: what is impossible in dif- fusive becomes feasible in a condensed state. Enzymes involved in the same metabolic pathway mutually interact and aggregate to build molecular machines in which the steps from initial reagents to ultimate products are conined in a separate ordered phase with respect to the cell bulk. On the other hand, highly organised cytoskeleton protein meshes allow for directed signalling across tissue microenvironment. The presence of an organised ‘interactome’ is the main prerequisite for metabolism to take place. Before going ahead, it is important to give a quantitative appre- ciation of the problem, building upon the impressive combina- torics of the interactome. In their 2011 paper entitled the ‘The Levinthal paradox of the interactome’ (Tompa and Rose, 2011), Peter Tompa and George Rose sketch a minimalistic combinato- rial model of the number of possible protein–protein interaction patterns. Taking yeast as a model organism, the authors estimate that approximately 4500 different proteins are expressed during log-phase growth, each present in 50 to more than 10 6 copies per cell (these numbers must be compared to the approximately 100 000 human protein species). These proteins have a median length of about 400 residues; assuming spherical shape and average density of 1.1 g cm -3 , the median protein would have a radius of 26.3 Å and a surface area of 8692 Å 2 . Next, considering an average protein–protein interface (PPI) of about 800 Å 2 , the equivalent of 22 interfacial residues and the usual Ramachandran admitted rotations around peptide bond, the authors estimate an average protein would have approximately 3540 distinguishable interfaces. Assuming the simplest (and extremely minimalistic) case of each of n proteins present in a single copy in the proteome with all proteins engaged in pairwise interactions, the total number of possible distinct patterns of interactions corresponds to N i = eLS © 2017, John Wiley & Sons, Ltd. www.els.net 1