Cyclotides are the largest family of cyclic proteins currently known, with more than 200 members isolated from several plant species of the Violaceae, Rubiaceae, Cucurbitaceae and Fabaceae plant families. Their amino and carboxyl ends are connected together with a peptide bond that forms their head-to-tail cyclic backbone. They contain approximately 30 amino acids, six of which are cysteine residues that are connected together with three disulfide bonds forming a cystine knot. The combination of the cyclic backbone and the knotted arrangement of three disulfide bonds, is termed a cyclic cystine knot (CCK) motif. Due to this unique topology, cyclotides can maintain their structure even when subjected to harsh thermal, chemical and enzymatic conditions. Their unique structural stability and sequence diversity make them attractive as pharmaceutical and agrochemical tools.
Cyclotides are classified into three subfamilies based on their structural features and sequence variations: the Möbius, bracelet, and trypsin inhibitor cyclotides. All cyclotides are gene encoded proteins and are synthesized from larger precursor proteins that comprise between one and three cyclotide domains. Cyclotide precursors in the Rubiaceae and Violaceae families have an endoplasmic reticulum (ER) signal peptide and one or more cyclotide domains each flanked by an N-terminal pro-peptide (NTR) and a C-terminal hydrophobic pro-peptide (TAIL).
The overall goal of this thesis was to attain a better understanding of the sequence diversity and biosynthetic processing mechanism of cyclotides by multi-disciplinary approaches. The first chapter provides a detailed review of the history and current knowledge of cyclotides, focussing on their distribution and evolution as well as their biosynthesis. This chapter provides the background necessary for the reader to understand gaps in current knowledge and therefore appreciate the rationale for studies described later in this thesis.
Chapter 2 describes the materials and methods used during this PhD project, with particular focus on the providing an overview of the techniques used to identify cyclotides. Variations to these generic procedures specific to individual experiments are described at the relevant place in later chapters.
Chapter 3 extends our knowledge on the sequence diversity and distribution of cyclotides. Twenty-six species from several plant families were screened for cyclotides at the peptide and cDNA levels. I discovered 14 novel cyclotides and one linear cyclotide from Viola kauaensis (Violaceae) and nine from Viola cornuta (Violaceae). I also discovered three linear cyclotides from Rinorea bengalensis (Violaceae). More than twenty partial precursor protein sequences were obtained as well. Overall, the study has provided a better understanding of the sequence diversity of cyclotides and increased the number of cyclotide-containing plants.
Chapter 4 focuses on the biosynthesis of a trypsin inhibitor cyclotide, MCoTI-II from Momordica cochinchinensis (Cucurbitaceae). I produced a partial precursor protein for MCoTI-II including the NTR region and cyclotide domain (NTR+ MCoTI-II) chemically by native chemical ligation of two separate chemically synthesised sections. Recent studies by others in my laboratory had shown that asparaginyl endopeptidase (AEP) is involved in processing the cyclotide kalata B1 from its precursor protein Oak1 at its proto-C-terminus. However, no enzyme has been identified that processes the N-terminal side of a cyclotide domain. The studies reported in Chapter 4 showed that recombinant human asparaginyl endopeptidase (AEP) cleaves the N-terminal side of MCoTI-II. The three dimensional structure of NTR+ MCoTI-II was determined. The NTR region was found to be disordered whereas the MCoTI-II region was found to be similar to the mature MCoTI-II.
Chapter 5 describes the isolation and sequencing of a hypothetical bracelet cyclotide (hcf-1 from Kadua centranthoides, Rubiaceae) which was predicted from an expressed sequence tag found in GenBank. The hcf-1 sequence was obtained by reduction, carbamidomethylation, enzyme digestion and tandem MS/MS. I also addressed the difficulties often encountered with in vitro oxidative folding specifically for bracelet cyclotides. The Möbius subfamily members have been proven to fold easily in vitro whereas bracelet subfamily members generally do not fold efficiently. A range of oxidative folding buffers were used to optimise the folding of the chemically synthesised hcf-1 including a mixture previously shown to enhance the folding of a bracelet cyclotide. However oxidative folding into the native conformation did not occur for hcf-1 indicating that the conditions previously established can not generally be applied. Furthermore, the folding of bracelet cyclotides is likely to involve chaperones in vivo given the poor folding in aqueous buffers.
Chapter 6 summarizes the overall aim of the thesis and the major findings obtained during this thesis. In summary, this thesis provides new information on the diversity and biosynthesis of cyclotides. These findings potentially have great importance in gaining a more complete understanding of the natural diversity of cyclotides and their biosynthetic processing mechanism.