I am generally inspired by research which may be applied in industry or for environmental management. During my Master Thesis in Portugal I used directed evolution techniques to engineer dye-decolourising peroxidases (DyPs) for lignin degradation. 

In my PhD project I work on bioreporter bacteria, which are genetically engineered to produce an easy measurable signal in response to a specific chemical compound. Bioreporters exploit sensory proteins, which upon binding of a target compound, can modify expression of a reporter protein. Current designs mostly use natural sensory proteins, but these cover only a limited range of detectable compounds.

 The goal of my project is to design sensory proteins with novel recognition specificities, based on the E. coli ribose binding protein (RbsB) as a starting point. I am using computational simulations (Rosetta) to predict variants, which instead of ribose may bind compounds with cyclic aliphatic and aromatic rings. I then produce and screen libraries of possible variants to isolate those with changed ligand binding properties, which I further characterize biochemically. Simulations and screenings are still a big challenge on a flexible protein such as RbsB, because the methods are not perfect and (for now) the gain of new ligand binding is small. I hope to be able to improve these methods so that ligand-binding design can become easier.



diogo.tavares fru@it unil.ch