Abstract:
As a product of millions of years of evolution, naturally occurring peptides display exquisite potency and selectivity towards protein targets due to their chemically similar binding surfaces. However, historically, peptides were not considered as drug-like compounds due to their poor oral bioavailability, membrane permeability and reduced metabolic stability. These disadvantages of linear peptides can be overcome by intramolecular cyclisation of the linear sequence to its corresponding cyclic peptide. Based on these favourable properties of cyclic peptides, this thesis describes the first total synthesis of eight naturally occurring cyclic peptides, namely versicoloritides A 1.18, B 1.19, and C 1.20 (chapter 2), tunicyclins C 1.22, and D 1.23 (chapter 3), dianthins G 1.24, and H 1.25 (chapter 4), and YM-280193 1.26 (chapter 5); as well as synthetic studies towards two related naturally occurring cyclic peptides, namely tunicyclin B 1.21 (chapter 3) and YM-254890 1.27 (chapter 5) (Figure i). In addition, to investigate the structure-activity-relationship pattern of the anabolic cyclic peptide dianthin G 1.24, six cyclic analogues 4.1-4.6 of dianthin G were also prepared (Figure ii). The synthetic strategy towards all ten cyclic peptides 1.18 to 1.27 in this doctoral project involved initial preparation of the respective linear precursors using Fmoc-solid phase peptide synthesis (Fmoc-SPPS). The linear precursors were then released from the acid-labile linker using TFA and subjected to the key head-to-tail cyclisation reaction using classical solution phase methods. This synthetic strategy was initially investigated with versicoloritides A 1.18, B 1.19 and C 1.20 (Scheme i). During the key head-to-tail cyclisation reaction of precursors 2.1-2.3, it was found that the undesired oligomerisation and epimerisation side-reactions could be reduced to acceptable levels by modifying the macrolactamisation reaction conditions. These optimised reaction conditions were thus used for the cyclisation of precursors 3.6-3.7 and 4.7- 4.8 towards the synthesis of tunicyclins C 1.22 and D 1.23, and dianthins G 1.24 and H 1.25, respectively (Scheme ii). In an analogous manner to the preparation of cyclic peptides 1.18-1.25, six cyclic dianthin G analogues 4.1-4.6 were also synthesised. Osteoblast proliferative assays of dianthins G 1.24 and H 1.25 and the six dianthin G analogues 4.1-4.6 then revealed that the alanine residues in analogues 4.3-4.6 could be further modified to improve the anabolic activity observed for dianthin G 1.24. iv Figure i. Structures of naturally occurring cyclic peptides that were aimed to be synthesised in this research project, namely versicoloritide A 1.18, versicoloritide B 1.19, versicoloritide C 1.20, tunicyclin B 1.21, tunicyclin C 1.22, tunicyclin D 1.23, dianthin G 1.24, dianthin H 1.25, YM-280193 1.26 and YM-254890 1.27. v Figure ii. Structures of dianthin G alanine-scan analogues 4.1-4.6, where the alanine substituted residues of dianthin G 1.24 are highlighted in red. Scheme i. Retrosynthetic strategy for versicoloritides A 1.18, B 1.19 and C 1.20. Scheme ii. Retrosynthetic strategy for tunicyclins C 1.22 and D 1.23, and dianthins G 1.24 and H 1.25. In addition to the cyclic peptides 1.18-1.25 and 4.1-4.6 mentioned above, the total synthesis of YM- 280193 1.26 is also described (Scheme iii). YM-280193 is an antiplatelet cyclic depsipeptide that consists of a highly N-methylated backbone, several depside bonds and predominantly unnatural residues such as β-hydroxyleucine and N-methylated dehydroalanine. The synthetic strategy towards YM-280193 1.26 involved initial preparation of the unnatural monomer and dipeptide fragments using conventional solution phase chemistry. These unnatural building blocks were then sequentially assembled using Fmoc-SPPS, after which a late-stage on-resin bis-alkylation-elimination reaction of the cysteine residue of 5.119 was employed to introduce the dehydroalanine moiety. The final macrolactamisation reaction of precursor 5.130 was then performed by solution phase methods to afford YM-280193 1.26. vii In addition to the cyclic peptides 1.18-1.25 and 4.1-4.6 mentioned above, the total synthesis of YM- 280193 1.26 is also described (Scheme iii). YM-280193 is an antiplatelet cyclic depsipeptide that consists of a highly N-methylated backbone, several depside bonds and predominantly unnatural residues such as β-hydroxyleucine and N-methylated dehydroalanine. The synthetic strategy towards YM-280193 1.26 involved initial preparation of the unnatural monomer and dipeptide fragments using conventional solution phase chemistry. These unnatural building blocks were then sequentially assembled using Fmoc-SPPS, after which a late-stage on-resin bis-alkylation-elimination reaction of the cysteine residue of 5.119 was employed to introduce the dehydroalanine moiety. The final macrolactamisation reaction of precursor 5.130 was then performed by solution phase methods to afford YM-280193 1.26. Scheme iii. Final retrosynthetic strategy for YM-280193 1.26.