NIRT: Combinatorial Engineering of Nanomachines: Building Novel Membrane Proteins via De Novo Design and Directed Evolution

University Of Pennsylvania, Philadelphia PA

Investigators

Abstract

This proposal was received in response to the announcement NSF 01-157. The overall objectives of this proposal are to: a) develop novel methods and tools for design and engineering of membrane proteins and protein assemblies based upon the integration of sophisticated computational chemistry techniques with in vitro directed molecular evolution; b) engineer novel membrane pores based upon the bacterial porin OmpF for controlling membrane vesicle permeability; c) engineer novel membrane fusion machines based upon influenza virus hemagglutinin for regulating bilayer fusion and membrane protein display; d) further the understanding of the physical and chemical properties underlying membrane protein structure and activity; e) train science and engineering students in these interdisciplinary nanoscience research methods. To incorporate stable proteins that span the ~4nm thickness of the lipid bilayer and that can mediate enzymatic functions or changes in conformation requires control of protein structure at the nanometer (and sub-nanometer) scale. In their proposal, novel membrane proteins will be constructed by combining rational design, partially random design via combinatorial libraries, and directed evolution. Two different systems will be focused on as starting points: bacterial porins, which are large permeability membrane pores, and the influenza virus protein hemagglutinin, which is a pH inducible membrane fusogen. A great deal of structural and functional data has been accumulated for both of these systems, and enormous potential exists for using them to build useful membrane based devices. Channels with altered and regulated permeability could be used to selectively deliver compounds to the ambient environment or selectively internalize and process external substrates. Similarly, gated fusogens could be used to control mixing between vesicles containing two different reactants and could also be engineered to act as switches that regulate the display of protein domains. Both gated pores and fusogens could also be incorporated into synthetic lipid assemblies in order to construct new "smart" materials, whose bulk elasticity and/or permeability are modulated in response to environmental signals. Beyond the specific utility of the proteins that they will engineer, the tools that they will develop and the new membrane proteins that emerge will provide valuable insight into the relatively primitive field of membrane protein design. The engineering of soluble proteins has burgeoned into an enormous field that is moving rapidly and is far ahead of the corresponding field for membrane proteins. In particular, the powerful tool of directed evolution, which has given rise to a wide range of new soluble proteins, has not been applied to the design of membrane-active proteins.