Computational design of artificial nuclear receptor proteins

In this project, we will use computer-based methods and experimental biology to design a novel nuclear receptor based biosensor, which may serve as an orthogonal signaling system for synthetic biology. This PhD thesis aims at the computational component of the project. This includes development of a computational protein design workflow to design a protein sensor for any given ligand (inducer) that, upon binding, is able to transfer the signal. Such a computationally designed switch is however important for synthetic biology where reprogramming and controlling a biological entity, such as yeast or plants is not only done to investigate regulatory pathways in the organisms but also to assign new characteristics or introduce a new behavior to the organism. For this project, we focus on inducers that are safe for the organism and the environment. Also, the computationally designed protein should belong to a type of signaling-pathway which is not naturally present in the organism and must allow for the binding event of the inducer to the ligand to lead to the activation or inactivation of a biochemical or genetic network. Nuclear receptors (NR) are the ideal protein switches for yeast and plants since they are not naturally present in these organisms. Therefore, they would not interfere with other established signaling pathways. Moreover, the modular structure of NRs allows to modify one main function of the domain and leave the functions of other domains generally unharmed. Therefore, we will focus during this project mainly on the steroid receptor family of the NR. Given the highly modular nature of NRs, the ability of the receptor to adapt to a completely different ligand, only requires the redesign of the LBD. The DNA binding domain and the amino-terminal domain are preserved at this stage of the project. The project is divided into several work packages combining computational and experimental work. The computational work packages can be summarized as below: 1. Selection of putative inducer ligands 2. Structural bioinformatics of the Nuclear Receptor Ligand-Binding Domain (LBD) 3. Computational Design of Novel Nuclear Receptor (XR) using Rosetta Modeling Suite 4. Optimization and Evaluation of XR activity The experimental work of the project will include the development of nuclear receptor assays and evaluation of the activity of putative inducers to find a suitable ligand for the design of artificial receptor. The computational design of XR will be followed by the bottom-up construction of NR signaling pathways, including co-activators in yeasts and plant cells. Finally, the optimization of the artificial receptor as an orthogonal signaling system for the survival of yeast cells.