Latent Thioesters in Protein Chemistry and Chemical Biology

The peptide thioester is an important tool in modern protein chemistry and chemical biology, particularly for the synthesis of proteins using a chemical reaction called Native Chemical Ligation (NCL). The synthetic thioester, which may contain unnatural amino acids or molecular probes, combines with additional peptide components to form a full length protein. NCL has enabled detailed studies of how proteins work and can provide access to pure samples of therapeutic peptides and proteins.

The driving force for NCL is the formation of the amide linkage. However we recently described a new transformation where peptides and proteins fragment to afford thioesters in a reverse-NCL type reaction, a process that occurs via Nitrogen to Sulfur (N to S) acyl transfer. This reaction yields the important thioester tools for NCL which can be extremely challenging to prepare. Peptides of synthetic or biological origin can serve as substrates simply by introduction of a C-terminal cysteine residue, which functions as a latent thioester. When peptides are additionally furnished with an N-terminal cysteine, head to tail cyclic peptides are produced through a retro NCL/NCL reaction sequence.

We have successfully applied the thioesters from our reaction to assembly of several synthetic bioactive peptides and semi-synthetic proteins. However, only through a detailed understanding of this fascinating transformation will we deliver a general, scalable, enabling methodology for peptide and protein synthesis, labelling, and chemical biology. Furthermore, this transformation is of fundamental interest since amide bonds, usually considered the more stable carboxylic acid derivatives, are broken under near physiological conditions in the absence of any enzymes, and formed in water, in the absence of typical peptide coupling reagents.

Our first goal is to improve the process through a detailed study in model peptides, exploring the ability of new terminal functional groups to expedite thioester formation. These experiments will ultimately tell us how reactions can be performed in shorter times, at room temperature or below. N to S acyl transfer can also be employed to remove a single amino acid from a proteins' N-terminus allowing access to N-terminal cysteine containing proteins which can also be very challenging to produce by other means. We will investigate optimal procedures for N-cysteinyl peptide production using fluorescence-based detection and apply an optimised protocol to an expressed protein. An interesting feature of our process is the ease with which head-to-tail cyclic peptides can be prepared. We propose that the cyclic product accumulates, in part, because solvent is excluded from the reaction site upon cyclisation and can explore this hypothesis using mass spectrometry in a model system. This process also highlights the dominance of NCL under our reaction conditions and additives that may temper competing NCL during thioester formation will also be explored.

An observed weakness is competing peptide hydrolysis at aspartate residues and so developing conditions, protecting groups, or aspartate "surrogates" that circumvent this problem will also be explored.

The second goal is to apply optimized protocols in more challenging contexts. First we will prepare an analogue of the HIV fusion inhibitor enfuvertide. Completion of the synthesis, in three sections, will provide a valuable proof of concept, employing two thioesters derived from our new methodology. The second application explores the synthesis of cyclic mirror image peptides derived from the naturally occurring beta defensin family of antimicrobials. The stability of these analogues is far superior to their L-peptide counterparts. To better understand the molecular basis for their antimicrobial activity we will conduct the synthesis of further analogues and prepare sufficient material for characterisation by NMR spectroscopy and X-ray crystallography.

In the not too distant past only biologists really had the ability to routinely manipulate the structure and function of proteins using the powerful tools of molecular biology. Chemists were quick to exploit a handful of selective bioconjugation reactions to further embellish expressed proteins with new functionality, natural and unnatural polymers, and create protein labelling and imaging technologies, new materials, and enhanced protein drugs. Recent developments have enabled the assembly of fully synthetic proteins that can also be highly novel with unique structures, and properties. This rich colour palette from which proteins can be washed forms the basis of a multi-billion dollar industry in peptide-based materials and pharmaceuticals. Indeed protein pharmaceuticals ("biologics") are set to dominate the market in the next five years.

The proposed research aims to better understand and apply a novel peptide and protein processing reaction discovered in our laboratory. It is potentially extremely valuable because it provides a straightforward route to the building blocks for protein synthesis using the Native Chemical Ligation (NCL) methodology and was, until demonstrated, considered impossible in the absence of additional protein domains called inteins. Indeed the only alternative route to producing the observed products is by the action of inteins. We believe that the proposed research has the ability to impact society in several ways:

academic impact: Our research, and the research it has stimulated will yield new developments and improvements in the production of e.g. peptide thioesters over the coming the years. Despite an obvious application to NCL the ease with which amide bonds can be cleaved and transformed hints at subtle gaps in our fundamental understanding of their chemistry and biology which has stimulated much interest on an international level. Although peptide chemistry is considered a "mature" science by some, the functionally complex context and how the environment influences protein reactivity is still poorly understood and the perceived "simplicity" of protein chemistry is surprising considering how chemists are regularly forced to rethink small molecule syntheses when a "trivial" reaction fails in a more complex setting. The organic chemistry of peptide bond synthesis is well developed has become a lucrative commercial pursuit though the departure of key practitioners into the commercial sector over the past couple of decades has arguably left the UK academic sector weaker in developers (though not users) of peptide and protein chemistry. Training obtained by pursuing this project proposal will equip a new investigator with the tools required to make new innovations in protein science.

economic impact: There are currently around 150 peptides and proteins on the market. Most therapeutic peptides are prepared by stepwise synthesis of shorter fragments that are then combined in convergent manner. Our proposed project is ideally placed to facilitate the production of such species and enable the preparation of protein thioesters that can be used to build therapeutic proteins or (fluorescently/isotopically/affinity) labelled proteins as probes and diagnostics. Furthermore, a novel method for producing cyclic peptides, a rapidly growing area of therapeutics will be developed, and in this project we will use mirror image antimicrobial peptides as a model system for cyclisation. Our methodology is unique since it alone has the potential to be directly applied to precursors of biological origin with no reliance on inteins. Reactions take place in water and do not require especially hazardous reagents. Ultimately, by illuminating our fundamental understanding, we can deliver a reliable and efficient protocol with application in protein synthesis and proteomics and in this context it should have high commercial potential.