Abstract
Optimizing synthetic biological systems, for example novel metabolic pathways, becomes more complicated with more protein components. One method of taming the complexity and allowing more rapid optimization is engineering external control into components. Pharmacology is essentially the science of controlling proteins using (mainly) small molecules, and a great deal of information, spread between different databases, is known about structural interactions between these ligands and their target proteins. In principle, protein engineers can use an inverse pharmacological approach to include drug response in their design, by identifying ligand-binding domains from natural proteins that are amenable to being included in a designed protein. In this context, "amenable" means that the ligand-binding domain is in a relatively self-contained subsequence of the parent protein, structurally independent of the rest of the molecule so that its function should be retained in another context. The SynPharm database is a tool, built on to the Guide to Pharmacology database and connected to various structural databases, to help protein engineers identify ligand-binding domains suitable for transfer. This article describes the tool, and illustrates its use in seeking candidate domains for transfer. It also briefly describes already-published proof-of-concept studies in which the CRISPR effectors Cas9 and Cpf1 were placed separately under the control of tamoxifen and mefipristone, by including ligand-binding domains of the Estrogen Receptor and Progesterone Receptor in modified versions of Cas9 and Cpf1. The advantages of drug control or the rival protein-control technology of optogenetics, for different purposes and in different situations, are also briefly discussed.