Science Literature

This article recently published in Accounts of Chemical Research discusses some key considerations for understanding how proteins bind in cellular machines, and offers a new approach to predicting binding. It provides some of the quantitative basis needed for putting irreducible complexity on a quantitative basis. Notably the authors seek to document the effect of mutations on reduced specificity of binding: "We have found that introducing mutations can significantly reduce specificity by introducing an additional binding mode." which reduction in specificity would be the first stage in the failure of a reducibly complex machine.

Worth reading both by those looking at the biochemistry and biology of protein machines and those pursuing the probability arguments of Intelligent Design.

Acc. Chem. Res., ASAP Article 10.1021 ar040204a S0001 4842(04)00204 3
Web Release Date: October 15, 2005
Copyright 2005 American Chemical Society
Mechanisms of Protein Assembly: Lessons from Minimalist Models
Yaakov Levy and Jose N. Onuchic
Center for Theoretical Biological Physics, Department of Physics, University of California at San Diego, 9500 Gilman Drive, La Jolla, California 92093

Received May 26, 2005

Abstract:
Many cellular functions rely on interactions among proteins and between proteins and nucleic acids. The limited success of binding predictions may suggest that the physical and chemical principles of protein binding have to be revisited to correctly capture the essence of protein recognition. In this Account, we discuss the power of reduced models to study the physics of protein assembly. Since energetic frustration is sufficiently small, native topology-based models, which correspond to perfectly unfrustrated energy landscapes, have shown that binding mechanisms are robust and governed primarily by the protein's native topology. These models impressively capture many of the binding characteristics found in experiments and highlight the fundamental role of flexibility in binding. The essential role of solvent molecules and electrostatic interactions in binding is also discussed. Despite the success of the minimally frustrated models to describe the dynamics and mechanisms of binding, the actual degree of frustration has to be explored to quantify the capacity of a protein to bind specifically to other proteins. We have found that introducing mutations can significantly reduce specificity by introducing an additional binding mode. Deciphering and quantifying the key ingredients for biological self-assembly is invaluable to reading out genomic sequences and understanding cellular interaction networks.

Introduction
The life of cells is orchestrated by a network of chemical reactions involving numerous proteins and nucleic acids and the transport of those molecules between cellular compartments. The remarkable efficiency of organizing these processes to yield a cellular function presents a major theoretical puzzle given the large number of molecular species and the crowded environment they inhabit. In the recent years, we have come to understand the assembly of the individual actors in this drama thanks to many cooperative efforts between experiments and theory. We now understand the main principles of folding kinetics[1,2] can often predict monomeric protein structure[3] and can even design novel protein structures.[4] However, knowing everything about isolated monomeric proteins does not give a complete understanding of function. Function requires change of structure and specific recognition to form large assemblies. These processes must be governed by the information stored in their sequences and structures. Furthermore, biomacromolecules are flexible with a rich repertoire of movements on various length and time scales. These motions are essential to determine the ability of a protein to bind different ligands at the same or different binding sites.[5,6] Deciphering the molecular and structural origins of high specificity as well as the catalytic promiscuity and multitasking of proteins is prerequisite for a quantitative understanding of the complexity and multidimensionality in genomes. This cooperation of many proteins and nucleic acids, which is largely "wireless", is quite intricate. Understanding the principles of biomolecular assembly in quantitative detail constitutes the basis for the molecular theory of biological networks.

Theoretical and computational studies of protein binding have concentrated on analyzing the structural and chemical properties of interfaces[7,8] as well as predicting the structure of the formed complexes and their binding affinity.[9,10] Understanding the organization of proteins into large complexes is required to understand their function and irreversible aggregation. The challenge of predicting the complex formed between pairs of proteins has been addressed for several years by docking two proteins using various models, which range from reduced models11[12] to atomistic ones[9] and include different flavors. Approaches to predict the structures of higher complexes, which are often defined as cellular machines, have been recently developed too. These approaches include, for example, combinatorial docking schemes[13] or fitting to cryo density maps at low resolution.[14] Some progress has been made in recent years in the performance of docking algorithms, yet their successes in predicting the structure of the protein complex are limited mainly to docking of the bound conformations of the complex subunits.

The inferiority of binding prediction to folding prediction is surprising because the conformational search required in binding processes of two folded proteins is smaller than that involved in protein folding. This shortcoming suggests that the physical and chemical principles of protein binding have to be revisited. The poor predictions of docking when using the conformations of the free subunits obviously indicates that protein flexibility is an important component in binding. Several docking approaches have introduced side-chain flexibility by using a rotamers library; however, it seems that backbone flexibility cannot be ignored.[15] It is likely, thus, that flexibility effects are still grossly underestimated as suggested from our recent association studies.[16-18] Solvent is also a critical component in protein association. While the protein cores are usually dry and contain a few water molecules, the interfaces in protein complexes are often very wet[19] (see Figure 1). Recently, it was found that a funneled potential for binding between proteins was obtained only upon solvation of assembly interfaces.[20] These observations provide a strong indication that water can be indispensable in protein assembly and undoubtedly in protein binding to DNA due to its highly charged surface.