Abstract
The recent progress made in protein design continues to pave the way for increasingly complex folds and functions. However, examples of rationally designed, de novo membrane proteins are few and far between. Here, the route towards a transmembrane β barrel assembled from designed β hairpins and templated by water-conducting $\alpha$-helical barrels is described. The α-helical barrels are redesigned to control the movement of water through their lumens; bioinformatic analysis of outer membrane β-barrel proteins directs design of a preliminary β-hairpin system towards a transmembrane pore; and, finally, membrane-active, β-strand peptides are designed for investigating the role of peptides as early transporters.α-Helical coiled coils have been used frequently as structural scaffolds, and they have been rationally designed to form open barrel structures with various oligomeric states. Adding polar residues to the open lumens of such barrels has been achieved, though with varied success. The first chapter of this thesis explores the effects of altering the stereochemistry of polar amino acids for the predictable and reliable hydration of hydrophobic α-barrel lumens. The X-ray crystal structures of several designs show a favourable positioning of hydroxyl side chains with the peptide backbone rather than water in the lumen. Threonine and allothreonine are interchanged to probe the level of control on the presence of water through the barrel lumen achievable by altering the position of the hydroxyl. In one peptide, this leads to alternate hexamer and antiparallel tetramer coiled-coil states, depending on side-chain stereochemistry, a phenomenon that has been observed previously in point mutants. Following this, allothreonine was also introduced into the α-helical, CCTM-VbIc peptide, which forms discrete pores in lipid bilayers, to determine the effects of this stereochemical change on the conductive properties.
In the second results chapter, a dataset of natural transmembrane β-barrel proteins is analysed to produce a series of amino-acid propensities for a variety of the environments within each protein. These propensities are applied to the rational design of β-hairpins by directing the choices of amino acid in each position within the intended transmembrane environment. From the thousands of sequences generated, the most suitable design for β-hairpins of strand lengths 6, 8, 10 and 12 amino acids are selected by filtering by key features of the β-hairpins in outer membrane proteins identified through analysis of the patterns across the dataset. These β-hairpin designs were looped to an α-barrel scaffold peptide and taken forward into solid phase peptide synthesis in the third chapter. The resulting β-hairpin:α-barrel peptides display a gradual loss of α helicity as the β-strand length increases. The 8-residue-strand β-hairpin displays annealing when heated to 95 °C, starting from an aggregate and forming a more α-helical structure. The designed constructs are compared to β hairpins from the α-hemolysin toxin.
The final results chapter describes a study of low-complexity, β-structured peptides and their interactions with lipid bilayers. A lysine-leucine repeat sequence was selected by contact analysis between nucleotide triphosphates (NTPs) and amino acids in NTP-binding proteins. Lysine was found to have the greatest propensity to be in contact with NTPs and was coupled with leucine, as a low propensity NTP binder, in an alternating repeat pattern to encourage both membrane interaction and β structure. These lysine-leucine (KL) repeat peptides have β-strand structure by circular dichroism and disrupt membranes in vesicle assays. Most importantly, KL peptides allow permeation of NTPs across POPC/POPG membranes to initiate RNA primer extension by ribozymes inside giant unilamellar vesicles.
| Date of Award | 28 Sept 2021 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Dek N Woolfson (Supervisor) & Chris L Willis (Supervisor) |