Self-assembling peptide nanopores are stabilized by a cooperative hydrogen-bond network

We have been using synthetic molecular evolution (generations of iterative library design and screening) to evolve large macromolecule-sized (5-10 nm diameter) “nanopores” that self-assemble into such controllable nanopores at low concentration. We identified the pHD peptides that self-assemble into nanopores in lipid bilayers at very low concentration, triggered by mildly acidic pH. We also evolved the closely related macrolittins, which have the same activity, but are not pH sensitive. Such peptide nanopore formation is unprecedented. These peptides fold into α-helices that, despite multiple charged and polar residues, insert into membrane-spanning configurations and stabilize the perimeter of large water-filled pores. Classical textbook concepts of protein folding in membranes do not predict this structure because the peptides appear to be too polar, overall. Since hydrophobicity alone does not account for their stability in membranes, these nanopores must also be stabilized by other interactions, such as H-bonds. Individual sidechain H-bonds in membrane proteins, or in contact with bulk water, are relatively weak. Yet, our atomistic molecular dynamics (MD) simulations show that charged and polar groups located along the polar surfaces of the nanopore forming peptides form dynamical, yet persistent, cooperative H-bond networks between peptides, lipids, and water that stabilize peptide nanopores. Such networks have been described for some membrane protein, but none are as extensive as the one we have identified in the pHD peptide and macrolittin nanopore structure. These nanopore forming peptides represent a new type of membrane protein structure.