Introduction next up previous contents
Next: The Cyclotides Up: Structural and evolutionary studies Previous: List of Figures


Naturally occurring circular proteins are becoming increasingly well known, with examples in bacteria, plants and animals discovered over recent years [1]. These topologically interesting proteins have a continuous cycle of peptide bonds in their backbone and, accordingly, are devoid of N- or C-termini. Such proteins were unknown a decade ago, and differ from previously known cyclic peptides, such as the immunosuppressant cyclosporin and other cyclic peptides found in micro-organisms, in that they are conventional gene products rather than the output of non-ribosomal synthetic processes. Furthermore, unlike non-ribosomally synthesised cyclic peptides, this new class of circular proteins are not limited to 5-12 amino acids but range in size from 14 to 78 amino acids and contain only conventionally coded amino acids without variations such as D-amino acids or N-methylation. As cyclic peptides seem to exhibit increased stability [2] they have aroused much interest as possible scaffolds for the engineering of protein based therapeutics.

Traditionally the use of peptides and proteins as drugs has met with limited success. Proteins tend to be unstable in biological systems, they often manifest poor bioavailability and are very costly to manufacture. Pharmaceutical research has generally focused on targets such as G-protein coupled receptors, ligand-gated ion channels and proteases which have as ligands small organic molecules that display desirable properties such as solubility and bioavailability [3]. On the other hand, proteins offer the potential for very high selectivity and may provide access to a range of macromolecular interactions that have been thought to be inaccessible for small organic molecules. Indeed, it has been pointed out that it is quite difficult to find small molecules that can affect many critical macromolecular interactions. Problems include the sheer size of chemical space, in which it has been estimated that approximately $ 10^{40}$ possible drug compounds may exist [4], and the need for extensive optimisation of hits from high throughput screening to provide therapeutic agents with the required pharmacokinetic properties [5].

Cyclisation of a protein backbone has the potential to remove some of the obstacles to developing effective protein based therapeutic agents, thus gaining access to their exquisite selectivity. Backbone cyclisation provides stabilisation relative to conventional proteins, both in a thermodynamic sense [6], and biologically [7] by removing at least one pathway to degradation, i.e., that of exopeptidase digestion. At least one class of cyclic peptides, the cyclotides, has been shown to be resistant to enzymatic activity and to a wide variety of adverse thermal and chemical conditions [2]. Furthermore, since the termini of conventional proteins are often flexible, and the degree of flexibility can be reduced by cyclisation, entropic factors can lead to improved receptor binding affinities of circular proteins over corresponding acyclic proteins, as may be the case in the cyclic trypsin inhibitor SFTI-I [8,9]. Thus, circular proteins, potentially, have a range of advantages over conventional proteins.

The broad aim of research in this field is to exploit the advantages of cyclisation and use these proteins and peptides as scaffolds for the engineering of novel therapeutics. In order to achieve this goal it is necessary to explore the limits of particular cyclic folds. Although backbone cyclisation offers the potential to stabilise proteins, as will be seen in the following sections, cyclic peptides exhibit a range of folds; some are rigid, some are flexible and although most display thermostability their resistance to a broad range of endoproteinases is variable. Presumably the structural characteristics of these proteins have evolved to best suit their particular function and clearly factors other than backbone cyclisation are important for stability and resistance to proteolytic digestion. Before these proteins and peptides can be utilised as scaffolds a clear understanding of their structural limitations is required as changes wrought through the introduction of novel motifs can disrupt the structural integrity of these proteins. For example amino acid substitutions can disrupt hydrogen bonding complexes that have been shown to stabilise some cyclic proteins, such as the trypsin inhibitor SFTI-1 [8,9]. Accordingly, it is vital that the limits of a particular cyclic fold be explored if it is to be successfully used as scaffold.

In addition to the engineering of novel moieties onto cyclic peptides the concept of backbone cyclisation may be utilised to improve the stability and, possibly, the bioavailability of a range of linear peptides with useful therapeutic activities. For this reason the determination of natural mechanisms of backbone cyclisation is of great interest. Although synthetic methods have been developed for the cyclisation of small peptides [10,12,11,13], the determination of natural mechanisms may enable the improvement of current technologies. Apart from one bacterial system the biosynthetic routes of most cyclic proteins is a mystery and a great deal of the work in this study deals with the elucidation of cyclisation in two types of plant cyclic peptides.

Related to the biosynthetic mechanism the evolution of cyclic peptides is of great interest. Cyclic peptides appear to be randomly distributed throughout the natural world, appearing in one species but not in closely related species. At this stage it is not possible to trace a thread of cyclic protein development through the history of evolution and whether this is because more cyclic proteins await discovery is an open question. At this stage the wide range of organisms that have cyclic peptides and the marked differences between these proteins would suggest that backbone cyclisation is the result of convergent evolution. One possible explanation for this could be the involvment of proteases in the cyclising process with evolution acting upon the precursors of linear proteins to provide the necessary conditions for cyclisation. This may explain the distribution of cyclic proteins in the natural world as relatively few mutations in existing proteins could bring about cyclisation without the need for the gradual evolution of a complex synthetic mechanism.

next up previous contents
Next: The Cyclotides Up: Structural and evolutionary studies Previous: List of Figures
Jason Mulvenna