Novel tools to map allosteric networks in proteins.

Proteins are a class of molecules that provides the majority of essential cellular functions such as catalysis, recognition and transport. These functions require binding of the protein to a 'target' molecule. Binding and the activity of the associated function are largely regulated by the concentration of the involved molecules. However, there are additional regulatory mechanisms in proteins; an important one is 'allostery', in which binding of an (additional) 'effector' molecule changes the activity of the protein towards the target molecule. Binding of target and effector molecule occurs at different sites on the protein surface. The allosteric effect can modulate (enhance/reduce) either target molecule binding or subsequent processing (e.g. enzymatic reaction). The detailed mechanism of allosteric regulation in proteins is largely unknown. We know that the information (signal) of the effector binding event has to cross the protein structure to affect the target site. A plausible assumption is that effector binding induces a 'chain' of conformational changes. The aim of the present project is to detect those pathways within protein structures that transmit the effector signal. Owing to the dynamic nature of proteins, we will run molecular simulations to sample a large number of protein conformations. We will analyse correlated conformational changes within the conformational ensembles to follow the signal transmission at the atomistic level.

Proteins can accurately and promptly respond to changes in the intra- and extra-cellular environment. This response can be modulated by an effector that modifies the property of the active site (substrate site) through binding to another region of the protein (effector site). By conveying information from the effector site to the substrate site, the protein can tune the response according to its location in space and time. This powerful mechanism, called allostery, plays an important role in many physiological processes and has recently provided promising targets for drug discovery. The use of allosteric modulators has shown a number of advantages over classic therapeutic agents targeting the substrate site. This has generated hopes for more selective treatments of some types of cancer, in particular for the Chronic Myeloid Leukaemia. We recently developed a method to describe allosteric transmission in a protein based on sampling the intrinsic flexibility by Molecular Dynamics, encoding structural changes with our Structural Alphabet and building a network model of local dynamics. The model assumes that information flows across the structure through the network by dynamic coupling, thereby conveying the actual signal transmission. Currently no such computational tool is available. The proposed project aims to develop the modelled software into a coherent package that will be provided as e-tools through a web interface. The method is based on the property of allosteric proteins to invoke correlated motions and conformational transitions of local structures (fragments) to generate an information flow across the structure. A data repository will hold the results of the test set of representative allosteric proteins and those of the Bcr-Abl protein. The repository will be open to contributions from other research groups, who will be able to 'publish' their results by uploading the data obtained from the server or the released software package.

Impact on Science The immediate impact of the proposed methodology will be on the analysis of single allosteric proteins. Computational and experimental biologists will have access to a comprehensive tool to model allosteric modulation in proteins of interest. An improved understanding of how protein function is modulated will impact the experimental design and inspire new strategies for pharmacological targeting. A later impact will be the ability to control protein function via allosteric modulation. It is predicted that understanding of allostery for key proteins will be helpful for researchers working on regulatory networks to progress in the comprehension of cellular regulation. Impact on Tools and Technology Biophysics and Computational Biology The core of the theoretical methodology is the combination of data on protein dynamics and a catalogue of local conformations (Structural Alphabet) to generate a network representation of dynamic coupling in a protein structure. This novel combination of methods from biophysics and computational biology will provide a set of powerful tools with unrestricted access. This will foster the dissemination and exploitation of the new technology. e-tools The computational methodologies, the software and the repository data will be provided by a web interface integrating open standards and the most recent web technologies. It is predicted that the users will profit from this centralised resource and the electronic delivery of tools and contents. Predictive Models The web interface will host a data repository for allosteric pathways represented in our network model. We predict that users will adopt our methodology and deposit their data in the repository. On a longer-term scale this collection of pathways could be mined for predictive purposes. On the short scale each network analysis will provide a map of signal transmission routes that could be targeted and tested by experimentally. Impact on Health The project proposal includes an important case of allosteric regulation that has been recently identified as relevant for pharmacological targeting: the oncogenic fusion protein Bcr-Abl. Bcr-Abl is the cause of Chronic Myeloid Leukaemia (CML) and one novel avenue to overcome the rise of resistant mutants is the combined use of orthosteric and allosteric drugs. The application of our method to this case will shed light on the mechanism behind this successful strategy. The impact of this insight will not be limited to Bcr-Abl, but it will extend the use of allosteric targeting to other kinases. Impact on Commerce Kinases are among the top drug target and the kinase-targeted drug market is predicted to grow substantially to over 50bn GBP by the end of 2010. Several pharmaceutical companies have developed drugs based on orthosteric inhibition. New strategies based on allosteric modulation can extend the scope of already marketed drugs or reduce the amount of ligands required addressing the issue of patient intolerance. Commercially interesting developments resulting from this project will be identified and exploited in co-operation with KCL Business. Impact on People One post-doctoral researcher will be trained in a highly rated scientific environment. He will profit from a interdisciplinary location with the opportunity to develop a network of contacts for further development of his scientific career. He will acquire new scientific knowledge and he will have the chance to develop a key core specialisation in computational analysis of allosteric modulation. This will constitute a central theme to achieve an independent position and to apply for independent fellowship career-development programmes. Impact on UK's Competitiveness The UK's position as a world leader biomedical and pharmaceutical research is based on innovative ideas and approaches. An automated computational analysis of allosteric modulation will be crucial in the design new more potent inhibitors.