Macmillan Research Group
Research Summary
Our most recent research has been aimed towards the further exploration of N→S acyl transfer reactions in native peptides and proteins (Scheme 1)
Scheme 1. Cysteine promoted C-terminal thioester formation.
This process was previously considered impossible in the absence of protein splicing elements called inteins, but can occur under relatively mild conditions. We now use it routinely to prepare thioester building blocks for NCL. Throughout our investigations (supported by EPSRC Project grant EP/J007560/1), we have sought to further understand and improve the process through exploration of new reagents and additives that can furnish additional C-terminal derivatives including acyl hydrazides, which can be employed advantageously in several instances.
Head-to-tail cyclic peptides can also form spontaneously by an NCL-type process when the peptide undergoing N→S acyl transfer also possesses an N-terminal cysteine residue. Our N→S acyl transfer method is unique in that it can be routinely applied to peptides of synthetic or recombinant origin, since it requires only the presence of a native amino acid to initiate thioester formation and cyclisation. It is amenable to bacterial protein expression, providing a valuable alternative and complementary route to intein-mediated splicing strategies. Recently we showed how bacterially expressed analogues of Sunflower Trypsin Inhibitor-1 (SFTI-1) could be readily transformed into cyclic peptides in sufficient quantities to enable structure validation and to develop a bioassay against related skin protease human Kallikrein 5 (hKLK5, Figure 1)
Figure 1. a) Overlay of the SFTI/Trypsin co-complex and KLK5 crystal structure emphasises their similarity around the SFTI-1 (magenta) binding site. b) Purification of a linear thioredoxin-SFTI fusion protein from E.coli. c) Mass spectrum of a cyclised, folded and purified SFTI-1 analogue. d) Exposure of cultured human keratinocytes to KLK5 in the presence of SFTI-I10G results in decreased Ca2+ influx, with the potential to mediate the inflammatory response.
In addition to producing head-to-tail cyclic peptides, we have recently found that this strategy is also potentially applicable to novel tail-to-sidechain topologies using synthetic branched amino acids (Figure 2). This topology is present in, for example, lariat peptides which have thus far not been accessible through organic synthesis. As well as potentially providing access to these challenging peptide structures, sidechain thioesters can lead to site-specifically modified peptides such as glycopeptides. As this process has no parallel in biology the longer term goal is to genetically encode the fully unprotected branched dipeptides, as well as additional cysteine analogues that have been shown to increase the rate of carboxyl activation in model systems.
Figure 2. A tail-to-sidechain cyclisation strategy employing a branched dipeptide in an unprotected linear precursor. b) A similar linkage is found inantimicrobial microcinJ25. c) Genetic screen for the incorportion on novel amino acids.
We have also pursued a strong interest in protein ligation methodologies, particularly Native Chemical Ligation (NCL) and its application to glycoprotein assembly. Realising that the challenges posed by particularly demanding biomolecules such as glycoproteins are unlikely to be met through total synthesis or molecular biology alone, our group has developed efficient routes for the introduction of synthetic saccharides and oligosaccharide mimics into synthetic and bacterially-derived peptides and proteins in a controlled manner. The semi-synthesis of a biologically active Erythropoietin analogue (left), was the result of one such collaborative project with the Kajihara group in Japan. We continue to apply NCL methodology to preparation of native glycoproteins using oligosaccharides derived from egg yolk.
Figure 3: Production of purified sialoglycopeptide (SGP) from dried egg yolk powder.
We have employed these enhanced processes in glycopeptide synthesis and remodelling (Figure 4).
Figure 4. a) Simply glycosylated semi-synthetic glycoproteins can be assembled through NCL and the sugar can be elaborated enzymatically following transfer of synthetic oligosaccharide oxazolines. b) Optimisation of Endo A production and examination of hydrolysis of the RNAse B glycan using the WT and hydrolytically inactive E173Q mutant. c) Glycosylated Glucagon-Like Peptide-1(GLP-1), produced using Endo A and oligosaccharide oxazolines. GLP-1 is a therapeutic peptide used to regulate insulin production.
