Current Protein and Peptide Science - Volume 10, Issue 2, 2009

Information is going to replace matter and energy as the primary stuff of the universe [1]. Information will provide a new basic unifying conceptual framework for describing and predicting physical events and reality in the 21st century [1]. However biological matter, from atomistic-molecular level to organ-organism level, constitutes the most sophisticated multiscale (in space and time) media in which information is codified, stored and communicated according to multilevel networks of interacting dynamical systems (biological information processing). Two of the most outstanding achievements of 20th century were the invention of computers and the birth of molecular biology. The advances made in these two fields over the past three decades have resulted not only in the generation of vast amounts of data and information, but also in a new understanding of the concept of information itself. Furthermore, modern science is unraveling the nature of information in numerous areas such as communication theory, biology, neuroscience, cognitive science, and biosemiotics, among others. Concurrently, computer based representations, simulations, modeling, and model-based reasoning are gaining in importance as a consequence of the increasing availability of computer power and sophisticated software for any kind of elaboration. In the concept of simulation as a model-based computational activity, the emphasis is on the generation and prediction of model behavior. Hence, simulation can be viewed as model-based knowledge-generation (computational experiments). Thus, simulation can be combined with other types of knowledge-generation techniques such as statistical elaboration, hypothesisprocessing and quantitative networks modeling with the aim of integrating the astonishing amount of the available biological data in a systems biology perspective. In the 20th century biologists successfully explained key functions and processes essential to the functioning of living beings, in terms of physical and chemical mechanisms ultimately driven by thermodynamic laws. These laws can be considered as universal constraints on the behavior of everything existing in the natural world. These restrictions are governed by two facts: in all natural processes energy is conserved and continuously degraded through the irreversible increase of entropy. The concepts and experimental methods of biochemistry and molecular biology succeeded in disclosing, in particular, numerous pathways and complex chains of chemical reactions involved in the metabolism and growth of cells and organisms. Many structural sequences of biomolecules have been mapped out in astonishing detail and have contributed to the relevant advances of current medicine and pharmacology. The main protagonists in these complicated series of events are the proteins. From a fixed repertoire of 20 amino acids, cells can manufacture an unlimited variety of proteins, with radically different shapes and properties, which govern molecular communication and information transfer. In this respect, the following definition of information due to von Baeyer [1] is particularly appropriate: “the word information derives from the Latin Informare (in+formare), meaning to give form, shape, or character to. It is therefore to be the formative principle of or to imbue with some specific character or quality”.

Proteins in their native folded states can possess multiple energy minima and they can show constant conformational fluctuations at physiological temperatures. In this article, we discuss the quantitative relationship between ligandinduced perturbation of such fluctuations, modeled as probability distributions of conformational substates, and allosteric coupling of ligand binding to different sites, as defined by linkage thermodynamics. We show that allosteric coupling between two binding events on the same protein is an inevitable consequence of ligand-induced perturbations of the probability distribution that represents conformational fluctuations in thermal equilibrium. When high resolution structural data of a protein in empty and ligand-bound forms are available, the COREX algorithm can provide, in principle, an excellent bridge between the energetics of substates distribution in the protein ensemble and structural coordinates. Here we propose a COREX-based strategic approach to link structural perturbations and the free energy changes of allosteric coupling. This strategy might be broadly useful in the endeavor of predicting how specific ligands allosterically regulate the function of specific proteins.

The phenomenon of intra-protein communication is fundamental to such processes as allostery and signaling, yet comparatively little is understood about its physical origins despite notable progress in recent years. This review introduces contemporary but distinct frameworks for understanding intra-protein communication by presenting both the ideas behind them and a discussion of their successes and shortcomings. The first framework holds that intra-protein communication is accomplished by the sequential mechanical linkage of residues spanning a gap between distal sites. According to the second framework, proteins are best viewed as ensembles of distinct structural microstates, the dynamical and thermodynamic properties of which contribute to the experimentally observable macroscale properties. Nuclear magnetic resonance (NMR) spectroscopy is a powerful method for studying intra-protein communication, and the insights into both frameworks it provides are presented through a discussion of numerous examples from the literature. Distinct from mechanical and thermodynamic considerations of intra-protein communication are recently applied graph and network theoretic analyses. These computational methods reduce complex three dimensional protein architectures to simple maps comprised of nodes (residues) connected by edges (inter-residue “interactions”). Analysis of these graphs yields a characterization of the protein's topology and network characteristics. These methods have shown proteins to be “small world” networks with moderately high local residue connectivities existing concurrently with a small but significant number of long range connectivities. However, experimental studies of the tantalizing idea that these putative long range interaction pathways facilitate one or several macroscopic protein characteristics are unfortunately lacking at present. This review concludes by comparing and contrasting the presented frameworks and methodologies for studying intra-protein communication and suggests a manner in which they can be brought to bear simultaneously to further enhance our understanding of this important fundamental phenomenon.

Allostery forms the basis of intra-molecular communications in various enzymes, however the underlying conformational changes are largely elusive. Recently, we have proposed to employ an elastic model based normal mode analysis to investigate the allosteric transitions in several molecular nanomachines (including myosin II, DNA polymerase and chaperonin GroEL). After combining with bioinformatics analysis of the evolutionary sequence variations, we have been able to identify the highly conserved and robust modes of collective motions that are capable of transmitting molecular signals over long distances.

Transducing environmental signals from the cell surface to the nucleus in order to evoke appropriate gene regulatory response requires an accurate and robust medium to propagate biological information. The structure of proteins and especially the dynamic properties of these structures allows for the uptake and restitution of biological information from and to the environment. To understand the functioning and regulation of signalling pathways we therefore have to understand how protein structures encode biological information. Towards this goal several computational methods have been carried out over the last years. First we will provide an overview of these in silico approaches. Next, using the well known SH2 domain as a case study, we describe two specific approaches in more detail to illustrate the similarities and differences between sequence-based and structure-based methods for the analysis of protein communication. Both methods address the same question yet from a different level of description. As a consequence both have their limits and a number of pros and cons that are discussed here. Together all the methods discussed here provide an arsenal of in silico approaches that may be used to understand how information content is maintained through protein structural dynamics, elucidating explicitly information transfer in signalling networks.

Communication within and across proteins is crucial for the biological functioning of proteins. Experiments such as mutational studies on proteins provide important information on the amino acids, which are crucial for their function. However, the protein structures are complex and it is unlikely that the entire responsibility of the function rests on only a few amino acids. A large fraction of the protein is expected to participate in its function at some level or other. Thus, it is relevant to consider the protein structures as a completely connected network and then deduce the properties, which are related to the global network features. In this direction, our laboratory has been engaged in representing the protein structure as a network of non-covalent connections and we have investigated a variety of problems in structural biology, such as the identification of functional and folding clusters, determinants of quaternary association and characterization of the network properties of protein structures. We have also addressed a few important issues related to protein dynamics, such as the process of oligomerization in multimers, mechanism of protein folding, and ligand induced communications (allosteric effect). In this review we highlight some of the investigations which we have carried out in the recent past. A review on protein structure graphs was presented earlier, in which the focus was on the graphs and graph spectral properties and their implementation in the study of protein structure graphs/networks (PSN). In this article, we briefly summarize the relevant parts of the methodology and the focus is on the advancement brought out in the understanding of protein structure-function relationships through structure networks. The investigations of structural/biological problems are divided into two parts, in which the first part deals with the analysis of PSNs based on static structures obtained from x-ray crystallography. The second part highlights the changes in the network, associated with biological functions, which are deduced from the network analysis on the structures obtained from molecular dynamics simulations.

The network paradigm is increasingly used to describe the dynamics of complex systems. Here we review the current results and propose future development areas in the assessment of perturbation waves, i.e. propagating structural changes in amino acid networks building individual protein molecules and in protein-protein interaction networks (interactomes). We assess the possibilities and critically review the initial attempts for the application of game theory to the often rather complicated process, when two protein molecules approach each other, mutually adjust their conformations via multiple communication steps and finally, bind to each other. We also summarize available data on the application of percolation theory for the prediction of amino acid network- and interactome-dynamics. Furthermore, we give an overview of the dissection of signals and noise in the cellular context of various perturbations. Finally, we propose possible applications of the reviewed methodologies in drug design.

Intramolecular and intermolecular communication is a privileged issue in G protein-Coupled Receptor (GPCR) function as the prominent role of these receptors is to respond to extracellular signals by catalyzing nucleotide exchange in intracellular G proteins. In the last decade or so we have applied much effort in elaborating computational strategies to infer the mechanisms of intramolecular and intermolecular communication in a number of GPCRs of the rhodopsin family. In this article, we review the most relevant achievements on the matter. In summary, the receptor sites of activating mutations or ligand-binding communicate with a common allosteric site in the cytosolic domains. This was inferred from the observation that local perturbations by activating mutations or ligands correlate with increases in solvent accessibility of the neighborhoods of the highly conserved E/DRY receptor motif. The latter turned out to be the primary recognition point for the C-terminus of the G protein α-subunit, independent of the receptor or the G protein type. In spite of the highly composite nature of the receptor-G protein interface, receptor contacts with the C-terminus of the α5-helix seem to be the major players in the receptor-catalyzed formation of a nucleotide exit route. The latter would lie in between the αF-helix and the β6/α5 loop, which detach from each other upon receptor binding, giving solvent accessibility to the nucleotide. A worthy inference of the studies is that GPCRs employ common pathways for the transfer of functionally relevant information.

Ligand-protein and protein-protein interactions play a pivotal role in any cellular process and function by means of complex and dynamic mechanisms that involve sophisticated intra- and intermolecular communication pathways. The deeper understanding of the molecular and structural mechanisms of these pathways of chemical information transfer constitutes the foundations of rational druggable target discovery and drug design. In this context the role of both molecular recognition/communication between the interacting partners and their quantitative/dynamic description constitute the crucial point. In this respect, many approaches at different level of complexity have been developed and applied to different druggable target like enzymes, membrane receptors and protein assembly. They mainly differ in the accuracy and resolution level of molecular description and, hence, in the derived quantitative molecular descriptors/predictors and ligand-target models. In this review, we will try to illustrate some selected examples of ligand-target receptor protein models, by comparatively considering both series of ligands (ligand-based communication modeling) and ligand-target complexes ( target-based communication modeling) in order to describe the relevant structural/dynamic features of chemical information transfer in the ligand/drug design endeavour.

This is with reference to the article entitled “Branched Peptides as Therapeutics”, by Alessandro Pini*, Chiara Falciani and Luisa Bracci, published in Current Protein and Peptide Science, Vol. 9, No. 5, October 2008, pp. 468-477. Due to an oversight the figures of this paper were interchanged on reproduction in the printing. The correct figures are now being presented below with their corresponding captions.