A real-time estimation of antibiotic in a bacterium is very crucial for understanding how that molecule is processed inside the cell. This will also help understand the mechanism of resistance. We propose to design a genetically encoded biosensor that can bind to the antibiotic of interest and produce a convenient signal. Using periplasmic binding protein (PBP) family, that is known to bind to diverse set of small molecules as a scaffold, we propose to engineer split PBP, consisting of globular domains, that can sandwich a small molecule of interest, an antibiotic in the present case. Earlier work with PBP used the whole protein with limited success; a single scaffold only gave restricted opportunity to carry mutagenesis and design. We also propose that two domains when undergo ligand mediated binding, will show predictable and large change in interdomain distance and orientation. The spatial change can be exploited for a quantitative FRET increase, which can be correlated to the concentration of the antibiotic. The proposal brings together all aspects of computational protein design, that include, loop modeling, stability and solubility enhancement, protein interface design and finally ligand docking. Our protein engineering approach carefully meshes computational modeling with flow cytomtery assisted high throughput screening to create functional proteins with applications in therapy and diagnostics.
A real-time estimation of antibiotic in a bacterium is very crucial for understanding how that molecule is processed inside the cell and will help understand mechanism of resistance in a pathogen. The goal of the project is to create a genetically encoded biosensor, capable of quantitative estimation of the molecule inside the cell. We propose to achieve the goal by exploiting large conformational change upon ligand mediated protein-protein interaction using a split protein system. The approach is generalizable towards other molecules of interest, for example biomarkers and metabolites.
