INTERACTION DESIGN · De novo design of affinity, specificity, and multispecificity in synthetic protein interaction networks

The two most fundamental features of molecular recognition are affinity and specificity. The proposed research aims to develop a new methodology for de novo design of small interaction networks not seen in Nature and to generate such networks with predetermined levels of specificity and affinity. In Preliminary Results we developed a methodology to de novo design protein binders of surfaces that have not been structurally characterized in bound form. We used this method to design three experimentally validated medium-affinity binders of the apo form of an acyl carrier protein (ACP), which has no co-crystal structures, demonstrating that interaction design could target a broad range of physicochemically appropriate surfaces. This synthetic network is the first multispecific system (one ACP interacts with three unrelated proteins), which is completely devoid of selective pressures for function other than binding, and will serve as a model to study how high affinity, specificity, and multispecificity might arise from non-interacting proteins. We will develop a multidisciplinary computational/experimental approach to substantially increase the affinity of the ACP to its binders by introducing stabilizing substitutions to residues on either one or both sides of each designed interface. Affinity will be further increased by experimental in vitro selection of higher-affinity variants of the ACP or of its partners. We will additionally erect specificity barriers between the ACP-binder pairs by redesigning the amino acid identities and backbones of interfacial loops. Specificity and affinity are important features in many protein-engineering applications, including the development of protein therapeutics and diagnostics, and the generation of synthetic signaling and metabolic networks. Small networks, such as the ones we produce here, may be future building blocks in these applications.