Computational Genomics of Signal Transduction

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Computational genomics of signal transduction Signal transduction is a universal biological process vital to all organisms. Due to their central role in disease, our own intrinsic signal transduction systems and those of bacterial pathogens are the primary targets of drug design. Our long-term goal is to understand how cells detect, transmit, and adapt to signals. The focus of this project is on the bacterial chemotaxis system, which is the best studied model for understanding fundamentals of signal transduction at the molecular level. This system resembles, both in complexity and mechanism, higher-order eukaryotic signal transduction systems and is a determinant of virulence in numerous bacterial pathogens. The chemotaxis signaling complex consists of chemoreceptors, an adaptor protein and a kinase. It is well-conserved and models for its overall structural organization have been produced. However, how signals are transmitted within a chemoreceptor and between chemoreceptor(s) and a kinase are poorly understood. We propose to build on our findings and capitalize on our tool development to specifically address the mechanisms underlying signaling by chemoreceptors and disentangle their diversity. We will perform a number of long all-atom molecular dynamics simulations to test the behavior of mutant chemoreceptors locked in signal-on and signal-off states. Biochemical, biophysical and behavioral characterization of these mutants will be carried out by our collaborators. We will also refine chemoreceptor:adaptor:kinase interfaces and contact sites in high-resolution crystal structures using evolutionary information. Tens of thousands of chemoreceptor sequences available in current databases are not only potential drug targets, but also an invaluable resource of evolutionary information needed to resolve ligand-binding and protein-protein interactions. However, extreme sequence variation and frequent events of gene loss and horizontal gene transfer impede their characterization. We aim at overcoming these barriers by carrying out comprehensive sequence/structure analyses of chemoreceptor sensory and signaling domains. We will classify microbial chemoreceptors and produce class-specific, high-quality hidden Markov models enabling their seamless identification in public databases. All this information will be integrated into MiST, a Microbial Signal Transduction database, which is freely available online and new domain models will be submitted to Pfam, the leading protein domain database. Current MiST capabilities will be enhanced with new search and download options and updated with a vast amount of sequences from the Human Microbiome Project and all other current metagenomics datasets, resulting in a resource that can better serve an even greater scientific community.